Compare commits

..
Author SHA1 Message Date
pedrooot 29ea00cab9 feat(aws): add single-parameter CloudFormation scan role 2026-09-21 13:05:37 +02:00
364 changed files with 2252 additions and 14162 deletions
+1 -10
View File
@@ -110,20 +110,11 @@ DJANGO_OUTPUT_S3_AWS_SECRET_ACCESS_KEY=""
DJANGO_OUTPUT_S3_AWS_SESSION_TOKEN=""
# The AWS region where your S3 bucket is located (e.g., "us-east-1")
# Required if the bucket uses SSE-KMS: download URLs are then signed with SigV4, which
# is scoped to this region, so it must match the bucket's
DJANGO_OUTPUT_S3_AWS_DEFAULT_REGION=""
# The name of the S3 bucket where scan output should be stored
DJANGO_OUTPUT_S3_AWS_OUTPUT_BUCKET=""
# The storage endpoint the API and Celery workers use to upload and list scan output
# (e.g. "http://minio:9000"). Leave empty on AWS S3. Set it when scan output is stored on
# S3-compatible object storage such as MinIO instead of real S3.
# If set without DJANGO_OUTPUT_S3_AWS_PUBLIC_ENDPOINT_URL below, report download URLs are
# signed against this internal host, and a browser outside the container network cannot open them.
DJANGO_OUTPUT_S3_AWS_ENDPOINT_URL=""
# The storage address the browser can reach, used only to sign report download URLs
# (e.g. "https://storage.example.com"). Leave empty on AWS S3. Set it when storage is
# only reachable inside the container network, such as MinIO on "http://minio:9000".
@@ -174,7 +165,7 @@ SENTRY_RELEASE=local
# REO_DEV_CLIENT_ID=
#### Prowler release version ####
NEXT_PUBLIC_PROWLER_RELEASE_VERSION=v5.44.1
NEXT_PUBLIC_PROWLER_RELEASE_VERSION=v5.43.0
# Social login credentials
SOCIAL_GOOGLE_OAUTH_CALLBACK_URL="${AUTH_URL}/api/auth/callback/google"
@@ -76,7 +76,7 @@ jobs:
### Changes
This PR updates the `OCI_COMMERCIAL_REGIONS` dictionary in `prowler/providers/oraclecloud/config.py` and the matching list in `ui/lib/provider-credentials/oci-regions.ts` with the latest regions fetched from the OCI Identity API (`list_regions()`).
This PR updates the `OCI_COMMERCIAL_REGIONS` dictionary in `prowler/providers/oraclecloud/config.py` with the latest regions fetched from the OCI Identity API (`list_regions()`).
- Government regions (`OCI_GOVERNMENT_REGIONS`) are preserved unchanged
- DOD regions (`OCI_US_DOD_REGIONS`) are preserved unchanged
+2 -3
View File
@@ -11,7 +11,6 @@ on:
- "v5.*"
paths:
- ".github/workflows/ui-e2e-tests-v2.yml"
- ".github/workflows/test-impact-analysis.yml"
- ".github/test-impact.yml"
- "ui/**"
- "api/**" # API changes can affect UI E2E
@@ -238,8 +237,8 @@ jobs:
- name: Add AWS credentials for testing
run: |
echo "AWS_ACCESS_KEY_ID=${E2E_AWS_PROVIDER_ACCESS_KEY}" >> .env
echo "AWS_SECRET_ACCESS_KEY=${E2E_AWS_PROVIDER_SECRET_KEY}" >> .env
echo "AWS_ACCESS_KEY_ID=${{ secrets.E2E_AWS_PROVIDER_ACCESS_KEY }}" >> .env
echo "AWS_SECRET_ACCESS_KEY=${{ secrets.E2E_AWS_PROVIDER_SECRET_KEY }}" >> .env
- name: Build API image from current code
# docker-compose.yml references prowlercloud/prowler-api:latest from the registry,
+1 -5
View File
@@ -68,7 +68,7 @@ vulnerabilities:
expired_at: 2026-11-30
# Declared in the SPDX manifest that ships inside PowerShell's MicrosoftTeams module
# (Modules/MicrosoftTeams/8.0.0/_manifest/spdx_2.2/manifest.spdx.json). Trivy reads that
# (Modules/MicrosoftTeams/7.9.0/_manifest/spdx_2.2/manifest.spdx.json). Trivy reads that
# SBOM and reports what it declares, which is not the same as what the image contains:
# there is no Node runtime and no node_modules anywhere in the image, and the .NET
# assemblies target net472, a Windows-only framework. Nothing here is reachable, and none
@@ -129,10 +129,6 @@ vulnerabilities:
purls:
- "pkg:npm/fast-uri"
expired_at: 2027-01-31
- id: CVE-2026-84292
purls:
- "pkg:npm/fast-uri"
expired_at: 2027-01-31
- id: CVE-2026-69192
purls:
- "pkg:npm/ip-address"
-16
View File
@@ -4,22 +4,6 @@ All notable changes to the **Prowler API** are documented in this file.
<!-- changelog: release notes start -->
## [1.44.0] (Prowler v5.43.0)
### 🐞 Fixed
- Report download URLs can be signed against a browser-reachable storage host via `DJANGO_OUTPUT_S3_AWS_PUBLIC_ENDPOINT_URL`, so downloads complete on deployments where storage is only reachable inside the container network [(#12552)](https://github.com/prowler-cloud/prowler/pull/12552)
- A scan report download no longer fails with a server error when `DJANGO_OUTPUT_S3_AWS_DEFAULT_REGION` is unset, which is common on storage with no meaningful region [(#12552)](https://github.com/prowler-cloud/prowler/pull/12552)
- Lapsed pending invitations are reported as expired and no longer block a new invitation for the same email [(#12831)](https://github.com/prowler-cloud/prowler/pull/12831)
### 🔐 Security
- `libsqlite3-0`, `gzip`, `perl-base` and `libpcre2-8-0` upgraded in the API container image, patching high Debian CVEs [(#12804)](https://github.com/prowler-cloud/prowler/pull/12804)
- PowerShell from 7.5.9 to 7.5.11 in the API container image, bundling .NET runtime 9.0.20 and patching CVE-2026-62901 [(#12811)](https://github.com/prowler-cloud/prowler/pull/12811)
- Bumped `anyio` to 4.14.2 to resolve CVE-2026-63374 [(#12848)](https://github.com/prowler-cloud/prowler/pull/12848)
---
## [1.43.0] (Prowler v5.42.0)
### 🔄 Changed
@@ -0,0 +1 @@
Bumped `anyio` to 4.14.2 to resolve CVE-2026-63374
@@ -0,0 +1 @@
`libsqlite3-0`, `gzip`, `perl-base` and `libpcre2-8-0` upgraded in the API container image, patching high Debian CVEs
@@ -0,0 +1 @@
PowerShell from 7.5.9 to 7.5.11 in the API container image, bundling .NET runtime 9.0.20 and patching CVE-2026-62901
@@ -1 +0,0 @@
API key authentication no longer locks the key row on every request and now throttles `last_used_at` updates to once per 60 seconds, preventing a hot key from serializing all its requests onto a single locked row
@@ -1 +0,0 @@
Adds a periodic sweep that drops orphaned Attack Paths temp Neo4j scan databases left behind when a worker or Neo4j crashes mid-scan, before they accumulate unbounded
@@ -1 +0,0 @@
Resources no longer keep a stale failed findings count forever when a scoped or imported scan for the same provider completes after a full scan, which used to make the full scan skip its own cleanup
@@ -0,0 +1 @@
Lapsed pending invitations are reported as expired and no longer block a new invitation for the same email
@@ -1 +0,0 @@
Providers whose most recent completed scan has no `completed_at` timestamp are no longer missing from every endpoint that reports a provider's latest scan, which now falls back to scan creation order instead of skipping the provider
@@ -1 +0,0 @@
Unify how every endpoint resolves a provider latest completed scan, so overlapping scans no longer make findings, compliance and mute rules read from different scans
@@ -1 +0,0 @@
OCI provider secrets keep the region as home region for credential validation and scans, instead of always using us-ashburn-1
@@ -1 +0,0 @@
Provider deletion no longer fails when the provider has Attack Paths scans recorded on a sink that is no longer configured, such as Neptune after moving back to Neo4j
-1
View File
@@ -1 +0,0 @@
`pyjwt` from 2.13.0 to 2.14.0, patching GHSA-ffc3-869f-jxw9
@@ -0,0 +1 @@
Report download URLs can be signed against a browser-reachable storage host via `DJANGO_OUTPUT_S3_AWS_PUBLIC_ENDPOINT_URL`, so downloads complete on deployments where storage is only reachable inside the container network
@@ -0,0 +1 @@
A scan report download no longer fails with a server error when `DJANGO_OUTPUT_S3_AWS_DEFAULT_REGION` is unset, which is common on storage with no meaningful region
@@ -1 +0,0 @@
Scan output uploads and downloads can now target S3-compatible object storage such as MinIO directly via `DJANGO_OUTPUT_S3_AWS_ENDPOINT_URL`, instead of relying on process-wide AWS environment variables that also hijacked unrelated AWS API calls
@@ -1 +0,0 @@
Scan report downloads from an S3 bucket with default SSE-KMS encryption no longer fail with an `InvalidArgument` error: when `DJANGO_OUTPUT_S3_AWS_DEFAULT_REGION` is set, presigned download URLs are signed with AWS Signature Version 4 for that region
@@ -1 +0,0 @@
`POST /api/v1/scans` again returns the new scan id in the response `task_args`, which had been empty since the scan broker publish moved to transaction commit
@@ -1 +0,0 @@
`DELETE /api/v1/tasks/{id}` requires the permission of the operation that queued the task and rejects provider deletions, and `GET /api/v1/tasks` hides tasks of providers outside the visibility of the role
@@ -1 +0,0 @@
Celery loggers are now declared explicitly in `custom_logging.py` so fatal worker errors are no longer silenced by `disable_existing_loggers=True`. All long-running services in `docker-compose.yml` now have `restart: unless-stopped` so containers recover automatically after unexpected crashes.
+6 -6
View File
@@ -45,7 +45,7 @@ dependencies = [
"gunicorn==26.0.0",
"uvloop==0.22.1",
"lxml==6.1.0",
"prowler @ git+https://github.com/prowler-cloud/prowler.git@v5.44",
"prowler @ git+https://github.com/prowler-cloud/prowler.git@master",
"psycopg2-binary==2.9.9",
"pytest-celery[redis] (==1.3.0)",
"sentry-sdk[django] (==2.56.0)",
@@ -71,7 +71,7 @@ name = "prowler-api"
package-mode = false
# Needed for the SDK compatibility
requires-python = ">=3.11,<3.13"
version = "1.45.1"
version = "1.44.0"
# Shared ruff baseline (kept in sync with mcp_server/pyproject.toml).
# target-version tracks this project's lowest supported Python.
@@ -375,7 +375,7 @@ constraint-dependencies = [
"pydantic-core==2.41.5",
"pygithub==2.8.0",
"pygments==2.20.0",
"pyjwt==2.14.0",
"pyjwt==2.13.0",
"pylint==3.2.5",
"pymsalruntime==0.18.1",
"pynacl==1.6.2",
@@ -476,8 +476,8 @@ constraint-dependencies = [
# to 1.9.10 until the SDK bump propagates to the pinned master rev.
#
# prowler@master hard-pins dulwich==0.23.0 and pyjwt==2.12.1 in [project.dependencies].
# dulwich 1.2.5 patches GHSA-897w-fcg9-f6xj (arbitrary file write) and pyjwt 2.14.0
# patches GHSA-ffc3-869f-jxw9 (HMAC/PEM key-confusion); a constraint cannot satisfy these
# dulwich 1.2.5 patches GHSA-897w-fcg9-f6xj (arbitrary file write) and pyjwt 2.13.0
# patches PYSEC-2026-179 (HMAC/JWK key-confusion); a constraint cannot satisfy these
# against the SDK's hard pins, so override them to the patched versions until the SDK
# bump propagates to the pinned master rev. pyjwt keeps the [crypto] extra because an
# override replaces the whole requirement; bare pyjwt would drop it from the consumers
@@ -500,5 +500,5 @@ override-dependencies = [
"microsoft-kiota-serialization-multipart==1.9.10",
"microsoft-kiota-serialization-text==1.9.10",
"dulwich==1.2.5",
"pyjwt[crypto]==2.14.0"
"pyjwt[crypto]==2.13.0"
]
@@ -207,11 +207,6 @@ def drop_database(database: str) -> None:
sink_module.get_backend().drop_database(database)
def list_databases() -> list[str]:
"""List database names on the ingest cluster. Temp scan DBs always live here."""
return ingest.list_databases()
def drop_subgraph(database: str, provider_id: str) -> int:
return sink_module.get_backend().drop_subgraph(database, provider_id)
@@ -13,7 +13,6 @@ from api.attack_paths.ingest.driver import (
get_session,
get_uri,
init_driver,
list_databases,
run_cypher,
)
@@ -26,6 +25,5 @@ __all__ = [
"get_session",
"get_uri",
"init_driver",
"list_databases",
"run_cypher",
]
@@ -165,14 +165,6 @@ def drop_database(database: str) -> None:
session.run(f"DROP DATABASE `{database}` IF EXISTS DESTROY DATA")
def list_databases() -> list[str]:
"""List every database name on the Neo4j temp-database cluster."""
# A cluster returns one row per hosting server, so dedupe on name
with get_session() as session:
result = session.run("SHOW DATABASES YIELD name RETURN DISTINCT name")
return [record["name"] for record in result]
def clear_cache(database: str) -> None:
"""Best-effort cache clear for a Neo4j database."""
from api.attack_paths.database import GraphDatabaseQueryException
+33 -45
View File
@@ -1,5 +1,4 @@
import logging
from datetime import timedelta
from math import isfinite
from uuid import UUID
@@ -7,7 +6,7 @@ from api.db_router import MainRouter
from api.models import TenantAPIKey, TenantAPIKeyManager
from cryptography.fernet import InvalidToken
from django.core.exceptions import ObjectDoesNotExist
from django.db.models import Q
from django.db import transaction
from django.utils import timezone
from drf_simple_apikey.backends import APIKeyAuthentication as BaseAPIKeyAuth
from drf_simple_apikey.crypto import get_crypto
@@ -19,15 +18,12 @@ from rest_framework_simplejwt.authentication import JWTAuthentication
logger = logging.getLogger(__name__)
# Writing on every request makes all requests of a busy key contend on one row
API_KEY_LAST_USED_AT_THROTTLE_SECONDS = 60
class OrphanedAPIKeyError(Exception):
"""Raised when an API key outlived the user that owns it.
The revocation is written by a plain `update()` before this is raised, so it is
already persisted by the time `authenticate` catches it and rejects the request.
Handled by `authenticate`, which commits the revocation written while detecting it
and then rejects the request with `AuthenticationFailed`.
"""
@@ -41,9 +37,8 @@ class TenantAPIKeyAuthentication(BaseAPIKeyAuth):
"""
Override to use admin connection, bypassing RLS during authentication.
Returns the validated API key row from a single read. `authenticate` builds
the auth claims from that same row instead of looking it up again, so a key
revoked or orphaned right after validation can't still authenticate.
Returns the validated API key row, locked with `select_for_update`, so callers
must run inside `transaction.atomic(using=MainRouter.admin_db)`.
"""
try:
payload = self.key_crypto.decrypt(key)
@@ -72,11 +67,9 @@ class TenantAPIKeyAuthentication(BaseAPIKeyAuth):
raise AuthenticationFailed("API Key has already expired.")
try:
# Loading `entity` in the same query keeps a user deleted after this read
# from turning the later `api_key.entity` access into a 500
api_key = (
self.model.objects.using(MainRouter.admin_db)
.select_related("entity")
.select_for_update()
.get(id=api_key_pk)
)
except ObjectDoesNotExist:
@@ -92,9 +85,8 @@ class TenantAPIKeyAuthentication(BaseAPIKeyAuth):
# Revoke it as well, so it stops showing up as active and later attempts fail
# the `revoked` check above like any other revoked key.
if api_key.entity_id is None:
self.model.objects.using(MainRouter.admin_db).filter(
id=api_key.id, revoked=False
).update(revoked=True)
api_key.revoked = True
api_key.save(update_fields=["revoked"], using=MainRouter.admin_db)
logger.warning(
"Revoked orphaned API key: prefix=%s tenant=%s",
api_key.prefix,
@@ -120,38 +112,34 @@ class TenantAPIKeyAuthentication(BaseAPIKeyAuth):
except ValueError:
raise AuthenticationFailed("Invalid API Key.")
try:
api_key = self._authenticate_credentials(request, key)
except OrphanedAPIKeyError:
raise AuthenticationFailed("No entity matching this api key.")
# Validation, the `last_used_at` update and the auth claims all read the same
# row, locked until the transaction ends. Looking the key up a second time to
# build the claims used to leave a window where a key revoked or orphaned right
# after passing validation still authenticated.
with transaction.atomic(using=MainRouter.admin_db):
try:
api_key = self._authenticate_credentials(request, key)
except OrphanedAPIKeyError:
# Rejected below instead of here: leaving the block normally commits
# the revocation `_authenticate_credentials` wrote, while raising from
# inside would roll it back.
pass
else:
# The prefix used to be checked by the second lookup
if api_key.prefix != prefix:
raise AuthenticationFailed("Invalid API Key.")
# The prefix used to be checked by the second lookup
if api_key.prefix != prefix:
raise AuthenticationFailed("Invalid API Key.")
api_key.last_used_at = timezone.now()
api_key.save(update_fields=["last_used_at"], using=MainRouter.admin_db)
self._throttled_touch_last_used_at(api_key)
entity = api_key.entity
return entity, {
"tenant_id": str(api_key.tenant_id),
"sub": str(entity.id),
"api_key_prefix": api_key.prefix,
}
entity = api_key.entity
return entity, {
"tenant_id": str(api_key.tenant_id),
"sub": str(entity.id),
"api_key_prefix": api_key.prefix,
}
@staticmethod
def _throttled_touch_last_used_at(api_key: TenantAPIKey) -> None:
"""Write `last_used_at` at most once per throttle interval, without locking the row."""
now = timezone.now()
stale_before = now - timedelta(seconds=API_KEY_LAST_USED_AT_THROTTLE_SECONDS)
if api_key.last_used_at is not None and api_key.last_used_at >= stale_before:
return
TenantAPIKey.objects.using(MainRouter.admin_db).filter(
id=api_key.id, revoked=False
).filter(
Q(last_used_at__isnull=True) | Q(last_used_at__lt=stale_before)
).update(last_used_at=now)
raise AuthenticationFailed("No entity matching this api key.")
class CombinedJWTOrAPIKeyAuthentication(BaseAuthentication):
+1 -1
View File
@@ -409,7 +409,7 @@ def batch_delete(tenant_id, queryset, batch_size=settings.DJANGO_DELETION_BATCH_
Args:
tenant_id (str): Tenant ID the queryset belongs to.
queryset: The queryset of objects to delete.
queryset (QuerySet): The queryset of objects to delete.
batch_size (int): The number of objects to delete in each batch.
Returns:
@@ -1,48 +0,0 @@
from django.db import migrations
TASK_NAME = "attack-paths-reap-orphaned-tmp-databases"
INTERVAL_HOURS = 6
def create_periodic_task(apps, schema_editor):
IntervalSchedule = apps.get_model("django_celery_beat", "IntervalSchedule")
PeriodicTask = apps.get_model("django_celery_beat", "PeriodicTask")
schedule, _ = IntervalSchedule.objects.get_or_create(
every=INTERVAL_HOURS,
period="hours",
)
PeriodicTask.objects.update_or_create(
name=TASK_NAME,
defaults={
"task": TASK_NAME,
"interval": schedule,
"enabled": True,
},
)
def delete_periodic_task(apps, schema_editor):
IntervalSchedule = apps.get_model("django_celery_beat", "IntervalSchedule")
PeriodicTask = apps.get_model("django_celery_beat", "PeriodicTask")
PeriodicTask.objects.filter(name=TASK_NAME).delete()
# Clean up the schedule if no other task references it
IntervalSchedule.objects.filter(
every=INTERVAL_HOURS,
period="hours",
periodictask__isnull=True,
).delete()
class Migration(migrations.Migration):
dependencies = [
("api", "0099_delete_tenant_onboarding_profile"),
("django_celery_beat", "0019_alter_periodictasks_options"),
]
operations = [
migrations.RunPython(create_periodic_task, delete_periodic_task),
]
+2 -65
View File
@@ -617,66 +617,9 @@ class Task(RowLevelSecurityProtectedModel):
resource_name = "tasks"
class ScanQuerySet(models.QuerySet):
"""Shared selectors for "the latest scan of a provider".
The queryset must already be scoped by the caller: manager, tenant, RBAC,
providers and database alias.
"""
# How "which completed scan is the provider's current one" is ordered.
LATEST_ORDER_BY = (
models.F("completed_at").desc(nulls_last=True),
models.F("inserted_at").desc(),
models.F("id").desc(),
)
def _eligible_for_latest(self) -> "ScanQuerySet":
"""Restrict to the scans that may be a provider's latest.
Returns:
ScanQuerySet: The completed scans.
"""
return self.filter(state=StateChoices.COMPLETED)
def latest_per_provider(self) -> "ScanQuerySet":
"""Pick each provider's latest scan with `DISTINCT ON (provider_id)`.
Returns:
ScanQuerySet: One scan per provider, the latest one.
"""
return (
self._eligible_for_latest()
.order_by("provider_id", *self.LATEST_ORDER_BY)
.distinct("provider_id")
)
def latest_ids_per_provider(self) -> list[UUID]:
"""Evaluate `latest_per_provider` and return the scan ids.
The ids are materialised so callers can pass them as a literal `IN`
list; as a subquery Postgres misestimates the row count and picks a
slow nested loop.
Returns:
list[UUID]: The id of each provider's latest scan.
"""
return list(self.latest_per_provider().values_list("id", flat=True))
def latest_first(self) -> "ScanQuerySet":
"""Order eligible scans newest first, without deduplicating per provider.
Expects the queryset to be already filtered to a single provider.
Returns:
ScanQuerySet: The eligible scans, latest first.
"""
return self._eligible_for_latest().order_by(*self.LATEST_ORDER_BY)
class Scan(RowLevelSecurityProtectedModel):
objects = ActiveProviderManager.from_queryset(ScanQuerySet)()
all_objects = ScanQuerySet.as_manager()
objects = ActiveProviderManager()
all_objects = models.Manager()
_SCOPING_SCANNER_ARG_KEYS_CACHE: tuple[str, ...] | None = None
@@ -783,12 +726,6 @@ class Scan(RowLevelSecurityProtectedModel):
name="scans_prov_state_ins_desc_idx",
),
# TODO This might replace `scans_prov_state_ins_desc_idx` completely. Review usage
# Since `ScanQuerySet`, no code path reads a provider's
# completed scans by `-inserted_at`. The only query left that
# matches this index (and `scans_prov_state_ins_desc_idx` above)
# is `GET /scans?filter[provider]=…&filter[state]=completed` with
# the default sort. Both are candidates to drop in a follow-up
# once production `pg_stat_user_indexes.idx_scan` confirms it.
models.Index(
fields=["tenant_id", "provider_id", "-inserted_at"],
condition=Q(state=StateChoices.COMPLETED),
+11 -66
View File
@@ -1,7 +1,7 @@
from enum import Enum
from api.db_router import MainRouter
from api.models import Integration, Provider, Role, Task, User
from api.models import Integration, Provider, Role, User
from django.db.models import Q, QuerySet
from rest_framework.exceptions import PermissionDenied
from rest_framework.permissions import BasePermission
@@ -17,50 +17,6 @@ class Permissions(Enum):
UNLIMITED_VISIBILITY = "unlimited_visibility"
# Revoking a task needs the permission of the operation that queued it.
# None and unmapped names are not revocable; a revoked provider deletion
# would leave the provider soft-deleted with nothing re-queuing the cleanup.
TASK_REVOKE_PERMISSIONS: dict[str, list[Permissions] | None] = {
"provider-connection-check": [Permissions.MANAGE_PROVIDERS],
"provider-deletion": None,
"integration-connection-check": [Permissions.MANAGE_INTEGRATIONS],
"integration-s3": [Permissions.MANAGE_INTEGRATIONS],
"integration-security-hub": [Permissions.MANAGE_INTEGRATIONS],
"integration-jira": [Permissions.MANAGE_INTEGRATIONS],
"scan-perform": [Permissions.MANAGE_SCANS],
"scan-perform-scheduled": [Permissions.MANAGE_SCANS],
"scan-compliance-overviews": [Permissions.MANAGE_SCANS],
"scan-compliance-reports": [Permissions.MANAGE_SCANS],
"scan-finding-group-summaries": [Permissions.MANAGE_SCANS],
"scan-report": [Permissions.MANAGE_SCANS],
"attack-paths-scan-perform": [Permissions.MANAGE_SCANS],
"findings-mute-latest-scans": [Permissions.MANAGE_SCANS],
"lighthouse-connection-check": [],
"lighthouse-provider-connection-check": [],
"lighthouse-provider-models-refresh": [],
}
def get_user_roles(user: User, tenant_id: str) -> list[Role]:
"""Return every role assigned to the user in the tenant."""
return list(
User.objects.using(MainRouter.admin_db)
.get(id=user.id)
.roles.using(MainRouter.admin_db)
.filter(tenant_id=tenant_id)
)
def roles_have_permissions(
roles: list[Role], required_permissions: list[Permissions]
) -> bool:
"""Return True when every required permission is granted by at least one role."""
return all(
any(getattr(role, permission.value, False) for role in roles)
for permission in required_permissions
)
class HasPermissions(BasePermission):
"""
Custom permission to check if the user's role has the required permissions.
@@ -78,11 +34,19 @@ class HasPermissions(BasePermission):
if not tenant_id:
return False
user_roles = get_user_roles(request.user, tenant_id)
user_roles = list(
User.objects.using(MainRouter.admin_db)
.get(id=request.user.id)
.roles.using(MainRouter.admin_db)
.filter(tenant_id=tenant_id)
)
if not user_roles:
return False
return roles_have_permissions(user_roles, required_permissions)
return all(
any(getattr(role, permission.value, False) for role in user_roles)
for permission in required_permissions
)
def get_role(user: User, tenant_id: str) -> Role:
@@ -121,25 +85,6 @@ def get_providers(role: Role) -> QuerySet[Provider]:
).distinct()
def get_tasks(role: Role) -> QuerySet[Task]:
"""Return the tasks visible to the role: tenant-wide ones and those of its providers."""
queryset = Task.objects.filter(tenant_id=role.tenant_id)
if role.unlimited_visibility:
return queryset
# Task has no provider FK, so match provider ids inside the stored kwargs.
# all_objects keeps a soft-deleted provider visible to its own groups, so the
# role that queued its deletion can still follow the task.
hidden = Q()
for provider_id in (
Provider.all_objects.filter(tenant_id=role.tenant_id)
.exclude(provider_groups__in=role.provider_groups.all())
.values_list("id", flat=True)
):
hidden |= Q(task_runner_task__task_kwargs__contains=str(provider_id))
return queryset.exclude(hidden) if hidden else queryset
def get_integrations(
role: Role, providers: QuerySet[Provider] | None = None
) -> QuerySet[Integration]:
+4 -9
View File
@@ -1,7 +1,7 @@
openapi: 3.0.3
info:
title: Prowler API
version: 1.45.1
version: 1.44.0
description: |-
Prowler API specification.
@@ -14823,9 +14823,7 @@ paths:
get:
operationId: api_v1_tasks_list
description: Retrieve a list of all tasks with options for filtering by name,
state, and other criteria. Tasks that reference a provider are only returned
when the role can access it; tasks without a provider reference are returned
for every role.
state, and other criteria.
summary: List all tasks
parameters:
- in: query
@@ -14924,8 +14922,7 @@ paths:
/api/v1/tasks/{id}:
get:
operationId: api_v1_tasks_retrieve
description: Fetch detailed information about a specific task by its ID. Tasks
tied to a provider outside the visibility of the role are not found.
description: Fetch detailed information about a specific task by its ID.
summary: Retrieve data from a specific task
parameters:
- in: query
@@ -14966,9 +14963,7 @@ paths:
delete:
operationId: api_v1_tasks_destroy
description: Try to revoke a task using its ID. Only tasks that are not yet
in progress can be revoked, and the caller needs the same permission as the
operation that queued the task (for example MANAGE_SCANS for a scan). Provider
deletions cannot be revoked.
in progress can be revoked.
summary: Revoke a task
parameters:
- in: path
@@ -1,9 +1,9 @@
import json
import time
from datetime import UTC, datetime, timedelta
from uuid import uuid4
import pytest
from api.authentication import API_KEY_LAST_USED_AT_THROTTLE_SECONDS
from api.db_router import MainRouter
from api.models import Membership, Role, TenantAPIKey, User, UserRoleRelationship
from api.signals import revoke_membership_api_keys, revoke_user_api_keys
@@ -11,7 +11,6 @@ from conftest import TEST_PASSWORD, get_api_tokens, get_authorization_header
from django.db.utils import ConnectionDoesNotExist
from django.urls import reverse
from drf_simple_apikey.crypto import get_crypto
from freezegun import freeze_time
from rest_framework.test import APIClient
from rest_framework_simplejwt.token_blacklist.models import (
BlacklistedToken,
@@ -528,7 +527,7 @@ class TestAPIKeyAuthentication:
def test_last_used_at_tracking(
self, create_test_user, tenants_fixture, api_keys_fixture
):
"""Verify last_used_at timestamp is set on first use and throttled after that."""
"""Verify last_used_at timestamp updates on each authentication."""
client = APIClient()
api_key = api_keys_fixture[0]
@@ -537,11 +536,7 @@ class TestAPIKeyAuthentication:
# Use API key to authenticate
api_key_headers = get_api_key_header(api_key._raw_key)
start = datetime.now(UTC)
with freeze_time(start):
first_response = client.get(
reverse("provider-list"), headers=api_key_headers
)
first_response = client.get(reverse("provider-list"), headers=api_key_headers)
assert first_response.status_code == 200
# Reload from database and check last_used_at is set
@@ -549,23 +544,17 @@ class TestAPIKeyAuthentication:
first_used_at = api_key.last_used_at
assert first_used_at is not None
# Using the same key again within the throttle interval does not rewrite it
# Use the same key again after a small delay
time.sleep(0.1)
second_response = client.get(reverse("provider-list"), headers=api_key_headers)
assert second_response.status_code == 200
# Reload and verify last_used_at was updated
api_key.refresh_from_db()
assert api_key.last_used_at == first_used_at
# Past the throttle interval, the next use refreshes it
later = start + timedelta(seconds=API_KEY_LAST_USED_AT_THROTTLE_SECONDS + 1)
with freeze_time(later):
third_response = client.get(
reverse("provider-list"), headers=api_key_headers
)
assert third_response.status_code == 200
api_key.refresh_from_db()
assert api_key.last_used_at > first_used_at
second_used_at = api_key.last_used_at
assert second_used_at is not None
assert second_used_at > first_used_at
@pytest.mark.django_db
@@ -1452,7 +1441,6 @@ class TestAPIKeyRLSBypass:
The update to last_used_at during authentication must also use the
admin database since it occurs before RLS context is established.
Past the throttle interval, using the key again refreshes the timestamp.
"""
client = APIClient()
api_key = api_keys_fixture[0]
@@ -1460,11 +1448,7 @@ class TestAPIKeyRLSBypass:
assert api_key.last_used_at is None
api_key_headers = get_api_key_header(api_key._raw_key)
start = datetime.now(UTC)
with freeze_time(start):
first_response = client.get(
reverse("provider-list"), headers=api_key_headers
)
first_response = client.get(reverse("provider-list"), headers=api_key_headers)
assert first_response.status_code == 200
@@ -1472,11 +1456,9 @@ class TestAPIKeyRLSBypass:
first_timestamp = api_key.last_used_at
assert first_timestamp is not None
later = start + timedelta(seconds=API_KEY_LAST_USED_AT_THROTTLE_SECONDS + 1)
with freeze_time(later):
second_response = client.get(
reverse("provider-list"), headers=api_key_headers
)
time.sleep(0.1)
second_response = client.get(reverse("provider-list"), headers=api_key_headers)
assert second_response.status_code == 200
api_key.refresh_from_db()
@@ -187,27 +187,6 @@ class TestRoutingByDatabasePrefix:
sink_backend_stub.drop_database.assert_called_once_with("db-tenant-abc")
mock_ingest.drop_database.assert_not_called()
def test_list_databases_always_routes_to_ingest(self, sink_backend_stub):
with patch("api.attack_paths.database.ingest") as mock_ingest:
mock_ingest.list_databases.return_value = ["db-tmp-scan-uuid-1"]
assert db_module.list_databases() == ["db-tmp-scan-uuid-1"]
mock_ingest.list_databases.assert_called_once_with()
def test_ingest_list_databases_dedupes_cluster_rows(self):
from api.attack_paths.ingest import driver as ingest_driver
with patch.object(ingest_driver, "get_session") as mock_get_session:
session = mock_get_session.return_value.__enter__.return_value
session.run.return_value = [{"name": "db-tmp-scan-uuid-1"}]
assert ingest_driver.list_databases() == ["db-tmp-scan-uuid-1"]
session.run.assert_called_once_with(
"SHOW DATABASES YIELD name RETURN DISTINCT name"
)
def test_clear_cache_routes_temp_to_ingest(self, sink_backend_stub):
with patch("api.attack_paths.database.ingest") as mock_ingest:
db_module.clear_cache("db-tmp-scan-uuid-1")
+24 -137
View File
@@ -5,18 +5,16 @@ from uuid import uuid4
import pytest
from api.authentication import (
API_KEY_LAST_USED_AT_THROTTLE_SECONDS,
OrphanedAPIKeyError,
SSEAuthentication,
TenantAPIKeyAuthentication,
)
from api.db_router import MainRouter
from api.models import TenantAPIKey, User
from api.models import TenantAPIKey
from django.db import connections
from django.db.models.query import QuerySet
from django.test import RequestFactory
from django.test.utils import CaptureQueriesContext
from freezegun import freeze_time
from rest_framework.exceptions import AuthenticationFailed
@@ -288,15 +286,14 @@ class TestTenantAPIKeyAuthentication:
assert str(exc_info.value.detail) == "This API Key has been revoked."
def test_authenticate_reads_the_api_key_once_without_a_row_lock(
def test_authenticate_reads_the_api_key_once_under_a_row_lock(
self, auth_backend, api_keys_fixture, request_factory
):
"""Test the API key is read a single time and no row is locked.
"""Test the API key is read a single time and the row is locked.
Validation, the `last_used_at` update and the claims must all come from the
same authoritative row: a second lookup would reopen the window where a key
revoked in between still authenticates. `SELECT ... FOR UPDATE` serialized
every request for a hot key onto one locked row and is not used any more.
same authoritative row: a second, unlocked lookup would reopen the window
where a key revoked in between still authenticates.
"""
api_key = api_keys_fixture[0]
@@ -313,40 +310,33 @@ class TestTenantAPIKeyAuthentication:
]
assert len(api_key_selects) == 1
assert "FOR UPDATE" not in api_key_selects[0]
assert "FOR UPDATE" in api_key_selects[0]
def test_authenticate_ignores_revocation_after_the_single_read(
def test_authenticate_ignores_revocation_after_the_locked_read(
self, auth_backend, api_keys_fixture, request_factory
):
"""Test the claims describe the row that was validated, not a later state.
Regression test: the key used to be looked up again to build the auth dict,
without rechecking `revoked` or `entity`. A key revoked or orphaned between
both reads still authenticated, and the claims came from that stale row.
There is now only a single read, so this race is closed by construction and
the revocation only takes effect on the next request.
both reads still authenticated, and the claims came from that stale row. With
a single locked read the write below cannot land mid-authentication, and the
revocation only takes effect on the next request.
"""
api_key = api_keys_fixture[0]
entity_at_validation = api_key.entity
original_authenticate_credentials = (
TenantAPIKeyAuthentication._authenticate_credentials
)
original_save = TenantAPIKey.save
def revoke_and_orphan_after_reading(self, request, key):
# Runs right after the single read `authenticate` will use to build the
# claims: the exact window a concurrent revocation used to slip into
result = original_authenticate_credentials(self, request, key)
def revoke_and_orphan_before_saving(instance, *args, **kwargs):
# Runs after validation, right before the claims are built: the exact
# window a concurrent revocation or user deletion used to slip into
TenantAPIKey.objects.filter(id=api_key.id).update(revoked=True, entity=None)
return result
return original_save(instance, *args, **kwargs)
request = request_factory.get("/")
request.META["HTTP_AUTHORIZATION"] = f"Api-Key {api_key._raw_key}"
with patch.object(
TenantAPIKeyAuthentication,
"_authenticate_credentials",
revoke_and_orphan_after_reading,
):
with patch.object(TenantAPIKey, "save", revoke_and_orphan_before_saving):
entity, auth_dict = auth_backend.authenticate(request)
assert entity == entity_at_validation
@@ -360,43 +350,6 @@ class TestTenantAPIKeyAuthentication:
assert str(exc_info.value.detail) == "This API Key has been revoked."
def test_authenticate_survives_owner_deleted_after_the_single_read(
self, auth_backend, api_keys_fixture, request_factory
):
"""Test a user deleted right after the read does not turn into a 500.
Without the row lock a concurrent user deletion can land between the read
and building the claims. `entity` is loaded by the same query, so no later
lookup can raise `DoesNotExist`.
"""
api_key = api_keys_fixture[0]
owner_id = api_key.entity_id
original_authenticate_credentials = (
TenantAPIKeyAuthentication._authenticate_credentials
)
def delete_owner_after_reading(self, request, key):
result = original_authenticate_credentials(self, request, key)
User.objects.using(MainRouter.admin_db).filter(id=owner_id).delete()
return result
request = request_factory.get("/")
request.META["HTTP_AUTHORIZATION"] = f"Api-Key {api_key._raw_key}"
with patch.object(
TenantAPIKeyAuthentication,
"_authenticate_credentials",
delete_owner_after_reading,
):
entity, auth_dict = auth_backend.authenticate(request)
assert auth_dict["sub"] == str(owner_id)
assert entity.id == owner_id
# From the next request on, the orphaned key is rejected with a 401
with pytest.raises(AuthenticationFailed):
auth_backend.authenticate(request)
def test_authenticate_expired_api_key(
self, auth_backend, create_test_user, tenants_fixture, request_factory
):
@@ -468,90 +421,24 @@ class TestTenantAPIKeyAuthentication:
if original_last_used:
assert api_key.last_used_at > original_last_used
def test_authenticate_updates_last_used_at_on_admin_database(
def test_authenticate_saves_to_admin_database(
self, auth_backend, api_keys_fixture, request_factory
):
"""Test that the `last_used_at` update runs against the admin database."""
"""Test that the API key save operation uses admin database."""
api_key = api_keys_fixture[0]
raw_key = api_key._raw_key
request = request_factory.get("/")
request.META["HTTP_AUTHORIZATION"] = f"Api-Key {raw_key}"
with CaptureQueriesContext(connections[MainRouter.admin_db]) as captured:
# Mock the save method to verify it's called with using='admin'
with patch.object(TenantAPIKey, "save") as mock_save:
auth_backend.authenticate(request)
api_key_updates = [
query["sql"]
for query in captured.captured_queries
if query["sql"].startswith("UPDATE") and '"api_keys"' in query["sql"]
]
assert len(api_key_updates) == 1
assert "last_used_at" in api_key_updates[0]
def test_authenticate_does_not_rewrite_last_used_at_within_throttle_interval(
self, auth_backend, api_keys_fixture, request_factory
):
"""Test that a second authentication within the throttle interval is a no-op write."""
api_key = api_keys_fixture[0]
raw_key = api_key._raw_key
request = request_factory.get("/")
request.META["HTTP_AUTHORIZATION"] = f"Api-Key {raw_key}"
# First call sets last_used_at
auth_backend.authenticate(request)
api_key.refresh_from_db()
first_used_at = api_key.last_used_at
assert first_used_at is not None
# Second call, still within the throttle interval, must issue no UPDATE
with CaptureQueriesContext(connections[MainRouter.admin_db]) as captured:
auth_backend.authenticate(request)
api_key_updates = [
query["sql"]
for query in captured.captured_queries
if query["sql"].startswith("UPDATE") and '"api_keys"' in query["sql"]
]
assert api_key_updates == []
api_key.refresh_from_db()
assert api_key.last_used_at == first_used_at
def test_authenticate_rewrites_last_used_at_after_throttle_interval(
self, auth_backend, api_keys_fixture, request_factory
):
"""Test that `last_used_at` is refreshed once it is older than the throttle interval."""
api_key = api_keys_fixture[0]
raw_key = api_key._raw_key
request = request_factory.get("/")
request.META["HTTP_AUTHORIZATION"] = f"Api-Key {raw_key}"
start = datetime.now(UTC)
with freeze_time(start):
auth_backend.authenticate(request)
api_key.refresh_from_db()
first_used_at = api_key.last_used_at
assert first_used_at is not None
later = start + timedelta(seconds=API_KEY_LAST_USED_AT_THROTTLE_SECONDS + 1)
with freeze_time(later):
with CaptureQueriesContext(connections[MainRouter.admin_db]) as captured:
auth_backend.authenticate(request)
api_key_updates = [
query["sql"]
for query in captured.captured_queries
if query["sql"].startswith("UPDATE") and '"api_keys"' in query["sql"]
]
assert len(api_key_updates) == 1
api_key.refresh_from_db()
assert api_key.last_used_at > first_used_at
# Verify save was called with using=admin_db
mock_save.assert_called_once_with(
update_fields=["last_used_at"], using=MainRouter.admin_db
)
def test_authenticate_returns_correct_auth_dict(
self, auth_backend, api_keys_fixture, request_factory
+1 -226
View File
@@ -1,16 +1,14 @@
from datetime import UTC, datetime, timedelta
from datetime import UTC, datetime
import pytest
from allauth.socialaccount.models import SocialApp
from api.db_router import MainRouter
from api.models import (
Provider,
ProviderComplianceScore,
Resource,
ResourceTag,
SAMLConfiguration,
SAMLDomainIndex,
Scan,
StateChoices,
StatusChoices,
TenantComplianceSummary,
@@ -526,226 +524,3 @@ class TestTenantComplianceSummaryModel:
assert summary1.id != summary2.id
assert summary1.requirements_passed != summary2.requirements_passed
def _latest_scan_fixture(tenant, provider, *, completed_at, inserted_at=None, **kwargs):
scan = Scan.objects.create(
tenant_id=tenant.id,
provider=provider,
trigger=kwargs.pop("trigger", Scan.TriggerChoices.MANUAL),
state=kwargs.pop("state", StateChoices.COMPLETED),
completed_at=completed_at,
**kwargs,
)
if inserted_at is not None:
# `inserted_at` is auto_now_add, so it has to be forced after the fact.
Scan.all_objects.filter(pk=scan.pk).update(inserted_at=inserted_at)
scan.refresh_from_db()
return scan
@pytest.mark.django_db
class TestScanQuerySetOrdering:
def test_completed_later_wins_over_inserted_later(
self, tenants_fixture, aws_provider
):
"""The scan that FINISHED last is current, not the one that started last."""
tenant, *_ = tenants_fixture
now = datetime.now(UTC)
finished_last = _latest_scan_fixture(
tenant,
aws_provider,
inserted_at=now - timedelta(hours=3),
completed_at=now,
)
_latest_scan_fixture(
tenant,
aws_provider,
inserted_at=now - timedelta(hours=1),
completed_at=now - timedelta(hours=1),
)
assert Scan.all_objects.filter(
tenant_id=tenant.id
).latest_ids_per_provider() == [finished_last.id]
def test_null_completed_at_provider_is_still_returned(
self, tenants_fixture, aws_provider
):
"""NULLS LAST, not `completed_at__isnull=False`.
Excluding NULL `completed_at` would drop the provider from every
"latest" endpoint instead of falling back to `inserted_at`.
"""
tenant, *_ = tenants_fixture
only_scan = _latest_scan_fixture(tenant, aws_provider, completed_at=None)
assert Scan.all_objects.filter(
tenant_id=tenant.id
).latest_ids_per_provider() == [only_scan.id]
def test_null_completed_at_never_outranks_a_finished_scan(
self, tenants_fixture, aws_provider
):
"""Postgres sorts NULLs first under DESC; NULLS LAST is what fixes it."""
tenant, *_ = tenants_fixture
now = datetime.now(UTC)
finished = _latest_scan_fixture(
tenant,
aws_provider,
inserted_at=now - timedelta(hours=2),
completed_at=now - timedelta(hours=2),
)
_latest_scan_fixture(
tenant,
aws_provider,
inserted_at=now,
completed_at=None,
)
assert Scan.all_objects.filter(
tenant_id=tenant.id
).latest_ids_per_provider() == [finished.id]
def test_id_breaks_an_exact_timestamp_tie_deterministically(
self, tenants_fixture, aws_provider
):
tenant, *_ = tenants_fixture
now = datetime.now(UTC)
scans = [
_latest_scan_fixture(
tenant, aws_provider, inserted_at=now, completed_at=now
)
for _ in range(3)
]
expected = max(scan.id for scan in scans)
picks = {
Scan.all_objects.filter(tenant_id=tenant.id).latest_ids_per_provider()[0]
for _ in range(5)
}
assert picks == {expected}
@pytest.mark.django_db
class TestScanQuerySetEligibility:
def test_unfinished_scans_are_excluded(self, tenants_fixture, aws_provider):
tenant, *_ = tenants_fixture
_latest_scan_fixture(
tenant,
aws_provider,
completed_at=None,
state=StateChoices.EXECUTING,
)
assert (
Scan.all_objects.filter(tenant_id=tenant.id).latest_ids_per_provider() == []
)
@pytest.mark.django_db
class TestScanQuerySetManagerChoice:
def test_active_manager_hides_soft_deleted_providers(
self, tenants_fixture, aws_provider
):
"""`Scan.objects` drops soft-deleted providers, `all_objects` keeps them.
The queryset must not decide this for the caller.
"""
tenant, *_ = tenants_fixture
scan = _latest_scan_fixture(
tenant, aws_provider, completed_at=datetime.now(UTC)
)
Provider.all_objects.filter(pk=aws_provider.pk).update(is_deleted=True)
assert Scan.all_objects.filter(
tenant_id=tenant.id
).latest_ids_per_provider() == [scan.id]
assert Scan.objects.filter(tenant_id=tenant.id).latest_ids_per_provider() == []
@pytest.mark.django_db
class TestScanQuerySetPerProviderScoping:
def test_one_scan_per_provider(self, tenants_fixture, aws_provider_pair):
tenant, *_ = tenants_fixture
provider_one, provider_two = aws_provider_pair
now = datetime.now(UTC)
newest_one = _latest_scan_fixture(tenant, provider_one, completed_at=now)
_latest_scan_fixture(tenant, provider_one, completed_at=now - timedelta(days=1))
newest_two = _latest_scan_fixture(tenant, provider_two, completed_at=now)
assert set(
Scan.all_objects.filter(tenant_id=tenant.id).latest_ids_per_provider()
) == {newest_one.id, newest_two.id}
def test_caller_filters_are_preserved(self, tenants_fixture, aws_provider_pair):
tenant, *_ = tenants_fixture
provider_one, provider_two = aws_provider_pair
now = datetime.now(UTC)
scan_one = _latest_scan_fixture(tenant, provider_one, completed_at=now)
_latest_scan_fixture(tenant, provider_two, completed_at=now)
assert Scan.all_objects.filter(
tenant_id=tenant.id, provider__in=[provider_one]
).latest_ids_per_provider() == [scan_one.id]
def test_latest_first_is_ordered_not_deduplicated(
self, tenants_fixture, aws_provider
):
tenant, *_ = tenants_fixture
now = datetime.now(UTC)
newest = _latest_scan_fixture(tenant, aws_provider, completed_at=now)
older = _latest_scan_fixture(
tenant, aws_provider, completed_at=now - timedelta(days=1)
)
ordered = list(
Scan.all_objects.filter(
tenant_id=tenant.id, provider_id=aws_provider.id
).latest_first()
)
assert [scan.id for scan in ordered] == [newest.id, older.id]
def test_tenant_isolation(self, tenants_fixture, aws_provider):
tenant, other_tenant, *_ = tenants_fixture
_latest_scan_fixture(tenant, aws_provider, completed_at=datetime.now(UTC))
assert (
Scan.all_objects.filter(tenant_id=other_tenant.id).latest_ids_per_provider()
== []
)
def test_empty_queryset_returns_empty_list(self, tenants_fixture):
tenant, *_ = tenants_fixture
assert (
Scan.all_objects.filter(tenant_id=tenant.id).latest_ids_per_provider() == []
)
@pytest.mark.django_db
class TestScanQuerySetPerProviderQuerysetShape:
def test_returns_a_queryset_not_a_list(self, tenants_fixture, aws_provider):
tenant, *_ = tenants_fixture
_latest_scan_fixture(tenant, aws_provider, completed_at=datetime.now(UTC))
qs = Scan.all_objects.filter(tenant_id=tenant.id).latest_per_provider()
# Callers chain .values(...) / .values_list(...) onto this.
assert qs.values_list("provider_id", flat=True).count() == 1
@pytest.mark.django_db
class TestScanQuerySetRelatedManager:
def test_reverse_relation_exposes_the_methods(self, tenants_fixture, aws_provider):
tenant, *_ = tenants_fixture
now = datetime.now(UTC)
newest = _latest_scan_fixture(tenant, aws_provider, completed_at=now)
_latest_scan_fixture(tenant, aws_provider, completed_at=now - timedelta(days=1))
assert aws_provider.scans.latest_first().first().id == newest.id
+24 -22
View File
@@ -215,34 +215,36 @@ class TestOracleCloudProviderSecret:
assert serializer.is_valid(), serializer.errors
assert "region" not in serializer.validated_data
def test_keeps_region_as_home_region(self):
serializer = OracleCloudProviderSecret(
data=self.valid_secret(region=" me-abudhabi-1 ")
)
def test_accepts_and_ignores_region_field(self):
secret = self.valid_secret(region="us-phoenix-1")
serializer = OracleCloudProviderSecret(data=secret)
assert serializer.is_valid(), serializer.errors
assert serializer.validated_data["region"] == "me-abudhabi-1"
def test_rejects_unknown_region(self):
serializer = OracleCloudProviderSecret(
data=self.valid_secret(region="mars-north-1")
)
assert not serializer.is_valid()
assert "region" in serializer.errors
@pytest.mark.parametrize("legacy_value", [None, "", {"name": "us-ashburn-1"}])
def test_drops_blank_or_non_string_region(self, legacy_value):
serializer = OracleCloudProviderSecret(
data=self.valid_secret(region=legacy_value)
)
assert serializer.is_valid(), serializer.errors
assert "region" not in serializer.validated_data
@pytest.mark.parametrize(
"legacy_field, legacy_value",
[
("region", None),
("region", ""),
("region", {"name": "us-ashburn-1"}),
],
)
def test_accepts_and_ignores_any_legacy_region_value(
self, legacy_field, legacy_value
):
serializer = OracleCloudProviderSecret(
data=self.valid_secret(**{legacy_field: legacy_value})
)
assert serializer.is_valid(), serializer.errors
assert legacy_field not in serializer.validated_data
class TestProviderSecretFieldSchema:
def test_oraclecloud_schema_region_is_not_deprecated(self):
def test_oraclecloud_schema_includes_legacy_region_field(self):
schema = ProviderSecretField._spectacular_annotation["field"]
oraclecloud_schema = next(
credential_schema
@@ -251,7 +253,7 @@ class TestProviderSecretFieldSchema:
== "Oracle Cloud Infrastructure (OCI) API Key Credentials"
)
assert "deprecated" not in oraclecloud_schema["properties"]["region"]
assert oraclecloud_schema["properties"]["region"]["deprecated"] is True
class TestKubernetesProviderSecret:
+8 -34
View File
@@ -172,7 +172,7 @@ class TestInitializeProwlerProvider:
)
@patch("api.utils.return_prowler_provider")
def test_initialize_oraclecloud_provider_passes_region_as_home_region(
def test_initialize_oraclecloud_provider_removes_region_string(
self, mock_return_prowler_provider
):
provider = MagicMock()
@@ -182,7 +182,7 @@ class TestInitializeProwlerProvider:
"fingerprint": "00:11:22:33:44:55:66:77",
"key_content": "fake-base64-key-content",
"tenancy": "ocid1.tenancy.oc1..fake",
"region": "me-abudhabi-1",
"region": "us-ashburn-1",
}
mock_return_prowler_provider.return_value = MagicMock()
@@ -193,7 +193,6 @@ class TestInitializeProwlerProvider:
fingerprint="00:11:22:33:44:55:66:77",
key_content="fake-base64-key-content",
tenancy="ocid1.tenancy.oc1..fake",
home_region="me-abudhabi-1",
)
@patch("api.utils.return_prowler_provider")
@@ -255,35 +254,11 @@ class TestProwlerProviderConnectionTest:
fingerprint="00:11:22:33:44:55:66:77",
key_content="fake-base64-key-content",
tenancy="ocid1.tenancy.oc1..aaaaaaaexample",
region=OraclecloudProvider._bootstrap_region,
provider_id="ocid1.tenancy.oc1..aaaaaaaexample",
raise_on_exception=False,
)
@patch("api.utils.return_prowler_provider")
def test_oraclecloud_connection_test_uses_stored_region_for_identity(
self, mock_return_prowler_provider
):
provider = MagicMock()
provider.uid = "ocid1.tenancy.oc1..aaaaaaaexample"
provider.provider = Provider.ProviderChoices.ORACLECLOUD.value
provider.secret.secret = {
"user": "ocid1.user.oc1..aaaaaaaexample",
"fingerprint": "00:11:22:33:44:55:66:77",
"key_content": "fake-base64-key-content",
"tenancy": "ocid1.tenancy.oc1..aaaaaaaexample",
"region": "me-abudhabi-1",
}
mock_return_prowler_provider.return_value = MagicMock()
prowler_provider_connection_test(provider)
mock_return_prowler_provider.return_value.test_connection.assert_called_once_with(
user="ocid1.user.oc1..aaaaaaaexample",
fingerprint="00:11:22:33:44:55:66:77",
key_content="fake-base64-key-content",
tenancy="ocid1.tenancy.oc1..aaaaaaaexample",
region="me-abudhabi-1",
region=getattr(
OraclecloudProvider,
"_bootstrap_region",
OraclecloudProvider._home_region,
),
provider_id="ocid1.tenancy.oc1..aaaaaaaexample",
raise_on_exception=False,
)
@@ -459,7 +434,7 @@ class TestGetProwlerProviderKwargs:
expected_result = {**secret_dict, **expected_extra_kwargs}
assert result == expected_result
def test_get_prowler_provider_kwargs_oraclecloud_maps_region_to_home_region(
def test_get_prowler_provider_kwargs_oraclecloud_removes_region(
self,
):
secret_dict = {
@@ -486,7 +461,6 @@ class TestGetProwlerProviderKwargs:
"key_content": "-----BEGIN PRIVATE KEY-----\nfake\n-----END PRIVATE KEY-----",
"tenancy": "ocid1.tenancy.oc1..fake",
"pass_phrase": "fake-passphrase",
"home_region": "us-ashburn-1",
}
def test_get_prowler_provider_kwargs_with_mutelist(self):
+6 -344
View File
@@ -60,7 +60,6 @@ from api.models import (
User,
UserRoleRelationship,
)
from api.rbac.permissions import TASK_REVOKE_PERMISSIONS
from api.rls import Tenant
from api.uuid_utils import datetime_to_uuid7
from api.v1.views import (
@@ -3363,7 +3362,7 @@ current-context: test-context
provider_secret = ProviderSecret.objects.get()
assert "region" not in provider_secret.secret
def test_provider_secrets_create_oraclecloud_stores_region(
def test_provider_secrets_create_oraclecloud_accepts_and_ignores_region(
self,
authenticated_client,
oraclecloud_provider,
@@ -3372,14 +3371,14 @@ current-context: test-context
authenticated_client,
oraclecloud_provider,
self._oraclecloud_secret(
key_content=" test-key-content ", region=" me-abudhabi-1 "
key_content=" test-key-content ", region=" us-ashburn-1 "
),
)
assert response.status_code == status.HTTP_201_CREATED
provider_secret = ProviderSecret.objects.get()
assert provider_secret.secret["key_content"] == "test-key-content"
assert provider_secret.secret["region"] == "me-abudhabi-1"
assert "region" not in provider_secret.secret
def test_provider_secrets_update_oraclecloud_without_region_stores_no_region(
self,
@@ -3412,7 +3411,7 @@ current-context: test-context
provider_secret.refresh_from_db()
assert "region" not in provider_secret.secret
def test_provider_secrets_update_oraclecloud_stores_region(
def test_provider_secrets_update_oraclecloud_accepts_and_ignores_region(
self,
authenticated_client,
oraclecloud_provider,
@@ -3430,7 +3429,7 @@ current-context: test-context
"type": "provider-secrets",
"id": str(provider_secret.id),
"attributes": {
"secret": self._oraclecloud_secret(region=" me-abudhabi-1 ")
"secret": self._oraclecloud_secret(region=" us-ashburn-1 ")
},
}
}
@@ -3443,7 +3442,7 @@ current-context: test-context
assert response.status_code == status.HTTP_200_OK
provider_secret.refresh_from_db()
assert provider_secret.secret["region"] == "me-abudhabi-1"
assert "region" not in provider_secret.secret
@pytest.mark.parametrize(
"attributes, error_code, error_pointer",
@@ -3952,43 +3951,6 @@ class TestScanViewSet:
mock_enqueue_scan_execution.assert_called_once()
# assert scan.scanner_args == expected_scanner_args
@patch("api.v1.views.enqueue_scan_execution_on_commit")
def test_scans_create_returns_the_scan_id_in_task_args(
self,
mock_enqueue_scan_execution,
authenticated_client,
okta_provider,
):
"""The 202 is a task, so `task_args` is the only place the scan id is.
It is serialized before the on_commit publish that would otherwise fill
the kwargs, so the record has to carry them from the start.
"""
payload = {
"data": {
"type": "scans",
"attributes": {"name": "New Scan"},
"relationships": {
"provider": {
"data": {"type": "providers", "id": str(okta_provider.id)}
}
},
}
}
response = authenticated_client.post(
reverse("scan-list"),
data=payload,
content_type=API_JSON_CONTENT_TYPE,
)
assert response.status_code == status.HTTP_202_ACCEPTED
scan = Scan.objects.get()
assert response.json()["data"]["attributes"]["task_args"] == {
"scan_id": str(scan.id),
"provider_id": str(okta_provider.id),
}
@patch("tasks.tasks.perform_scan_task.apply_async")
def test_scans_create_queues_scan_when_provider_has_active_scan(
self,
@@ -5240,7 +5202,6 @@ class TestTaskViewSet:
@patch("api.v1.views.AsyncResult", return_value=Mock())
def test_tasks_revoke(self, mock_async_result, authenticated_client, tasks_fixture):
_, task2 = tasks_fixture
self._set_task_name(task2, "scan-perform")
response = authenticated_client.delete(
reverse("task-detail", kwargs={"pk": task2.id})
)
@@ -5256,311 +5217,12 @@ class TestTaskViewSet:
def test_tasks_revoke_invalid_status(self, authenticated_client, tasks_fixture):
task1, _ = tasks_fixture
self._set_task_name(task1, "scan-perform")
response = authenticated_client.delete(
reverse("task-detail", kwargs={"pk": task1.id})
)
# Task status is SUCCESS
assert response.status_code == status.HTTP_400_BAD_REQUEST
@staticmethod
def _set_task_name(task, name):
task.task_runner_task.task_name = name
task.task_runner_task.save(update_fields=["task_name"])
@staticmethod
def _set_task_kwargs(task, kwargs):
task.task_runner_task.task_kwargs = json.dumps(repr(kwargs))
task.task_runner_task.save(update_fields=["task_kwargs"])
@staticmethod
def _client_with_role(tenant, factory, **permissions):
user = User.objects.create_user(
name=f"revoker-{uuid4()}",
email=f"revoker-{uuid4()}@prowler.com",
password=TEST_PASSWORD,
)
Membership.objects.create(
user=user, tenant=tenant, role=Membership.RoleChoices.MEMBER
)
flags = {
"manage_users": False,
"manage_account": False,
"manage_billing": False,
"manage_providers": False,
"manage_integrations": False,
"manage_scans": False,
"unlimited_visibility": True,
**permissions,
}
role = Role.objects.create(
name=f"revoker-{uuid4()}", tenant_id=tenant.id, **flags
)
UserRoleRelationship.objects.create(user=user, role=role, tenant_id=tenant.id)
return factory(user, tenant)
@patch("api.v1.views.AsyncResult")
def test_tasks_revoke_without_permission_is_forbidden(
self, mock_async_result, authenticated_client_no_permissions_rbac, tasks_fixture
):
_, pending_task = tasks_fixture
self._set_task_name(pending_task, "provider-connection-check")
response = authenticated_client_no_permissions_rbac.delete(
reverse("task-detail", kwargs={"pk": pending_task.id})
)
assert response.status_code == status.HTTP_403_FORBIDDEN
mock_async_result.return_value.revoke.assert_not_called()
@pytest.mark.parametrize(
"task_name, permissions, expected_status",
[
(
"provider-connection-check",
{"manage_providers": True},
status.HTTP_202_ACCEPTED,
),
(
"provider-connection-check",
{"manage_scans": True},
status.HTTP_403_FORBIDDEN,
),
("scan-perform", {"manage_scans": True}, status.HTTP_202_ACCEPTED),
(
"scan-perform-scheduled",
{"manage_providers": True},
status.HTTP_403_FORBIDDEN,
),
(
"integration-jira",
{"manage_integrations": True},
status.HTTP_202_ACCEPTED,
),
("integration-jira", {"manage_providers": True}, status.HTTP_403_FORBIDDEN),
("lighthouse-connection-check", {}, status.HTTP_202_ACCEPTED),
],
)
@patch("api.v1.views.AsyncResult")
def test_tasks_revoke_requires_originating_operation_permission(
self,
mock_async_result,
authenticated_client_for_tenant_factory,
tenants_fixture,
tasks_fixture,
task_name,
permissions,
expected_status,
):
tenant, *_ = tenants_fixture
_, pending_task = tasks_fixture
self._set_task_name(pending_task, task_name)
client = self._client_with_role(
tenant, authenticated_client_for_tenant_factory, **permissions
)
response = client.delete(reverse("task-detail", kwargs={"pk": pending_task.id}))
assert response.status_code == expected_status
if expected_status == status.HTTP_202_ACCEPTED:
mock_async_result.return_value.revoke.assert_called_once()
else:
mock_async_result.return_value.revoke.assert_not_called()
@patch("api.v1.views.AsyncResult")
def test_tasks_revoke_provider_deletion_is_forbidden_even_for_admin(
self, mock_async_result, authenticated_client, tasks_fixture
):
_, pending_task = tasks_fixture
self._set_task_name(pending_task, "provider-deletion")
response = authenticated_client.delete(
reverse("task-detail", kwargs={"pk": pending_task.id})
)
assert response.status_code == status.HTTP_403_FORBIDDEN
mock_async_result.return_value.revoke.assert_not_called()
@patch("api.v1.views.AsyncResult")
def test_tasks_revoke_unmapped_task_is_forbidden(
self, mock_async_result, authenticated_client, tasks_fixture
):
_, pending_task = tasks_fixture
assert pending_task.task_runner_task.task_name not in TASK_REVOKE_PERMISSIONS
response = authenticated_client.delete(
reverse("task-detail", kwargs={"pk": pending_task.id})
)
assert response.status_code == status.HTTP_403_FORBIDDEN
mock_async_result.return_value.revoke.assert_not_called()
def test_every_rls_task_has_revoke_permissions(self):
from config.celery import RLSTask, celery_app
rls_task_names = {
name for name, task in celery_app.tasks.items() if isinstance(task, RLSTask)
}
assert rls_task_names
assert rls_task_names <= set(TASK_REVOKE_PERMISSIONS)
@patch("api.v1.views.AsyncResult")
def test_tasks_hidden_for_providers_outside_role_visibility(
self,
mock_async_result,
authenticated_client_no_permissions_rbac,
tasks_fixture,
aws_provider_pair,
):
client = authenticated_client_no_permissions_rbac
limited_user = client.user
tenant = Membership.objects.filter(user=limited_user).first().tenant
allowed_provider, denied_provider = aws_provider_pair
allowed_task, denied_task = tasks_fixture
self._set_task_kwargs(
allowed_task,
{"tenant_id": str(tenant.id), "provider_id": str(allowed_provider.id)},
)
self._set_task_name(denied_task, "provider-deletion")
self._set_task_kwargs(
denied_task,
{"tenant_id": str(tenant.id), "provider_id": str(denied_provider.id)},
)
provider_group = ProviderGroup.objects.create(
name="limited-task-group", tenant_id=tenant.id
)
ProviderGroupMembership.objects.create(
tenant_id=tenant.id,
provider_group=provider_group,
provider=allowed_provider,
)
RoleProviderGroupRelationship.objects.create(
tenant_id=tenant.id,
role=limited_user.roles.first(),
provider_group=provider_group,
)
response = client.get(reverse("task-list"))
assert response.status_code == status.HTTP_200_OK
assert [item["id"] for item in response.json()["data"]] == [
str(allowed_task.id)
]
response = client.get(reverse("task-detail", kwargs={"pk": denied_task.id}))
assert response.status_code == status.HTTP_404_NOT_FOUND
response = client.delete(reverse("task-detail", kwargs={"pk": denied_task.id}))
assert response.status_code == status.HTTP_404_NOT_FOUND
mock_async_result.return_value.revoke.assert_not_called()
@patch("api.v1.views.AsyncResult")
def test_tasks_of_soft_deleted_provider_stay_visible_to_its_groups(
self,
mock_async_result,
authenticated_client_for_tenant_factory,
tenants_fixture,
tasks_fixture,
aws_provider_pair,
):
tenant, *_ = tenants_fixture
provider, _ = aws_provider_pair
finished_task, pending_task = tasks_fixture
client = self._client_with_role(
tenant,
authenticated_client_for_tenant_factory,
manage_providers=True,
unlimited_visibility=False,
)
provider_group = ProviderGroup.objects.create(
name="own-group", tenant_id=tenant.id
)
ProviderGroupMembership.objects.create(
tenant_id=tenant.id, provider_group=provider_group, provider=provider
)
RoleProviderGroupRelationship.objects.create(
tenant_id=tenant.id,
role=client.user.roles.first(),
provider_group=provider_group,
)
for task, name in (
(finished_task, "provider-deletion"),
(pending_task, "provider-connection-check"),
):
self._set_task_name(task, name)
self._set_task_kwargs(
task, {"tenant_id": str(tenant.id), "provider_id": str(provider.id)}
)
provider.is_deleted = True
provider.save()
response = client.get(reverse("task-detail", kwargs={"pk": finished_task.id}))
assert response.status_code == status.HTTP_200_OK
response = client.delete(reverse("task-detail", kwargs={"pk": pending_task.id}))
assert response.status_code == status.HTTP_202_ACCEPTED
mock_async_result.return_value.revoke.assert_called_once()
def test_tasks_without_provider_stay_visible_for_limited_roles(
self, authenticated_client_no_permissions_rbac, tasks_fixture, aws_provider_pair
):
response = authenticated_client_no_permissions_rbac.get(reverse("task-list"))
assert response.status_code == status.HTTP_200_OK
assert len(response.json()["data"]) == len(tasks_fixture)
def test_tasks_list_without_role_is_forbidden(
self, authenticated_client_rbac_noroles, tasks_fixture
):
response = authenticated_client_rbac_noroles.get(reverse("task-list"))
assert response.status_code == status.HTTP_403_FORBIDDEN
def test_tasks_revoke_without_permission_hides_task_status(
self, authenticated_client_no_permissions_rbac, tasks_fixture
):
finished_task, _ = tasks_fixture
self._set_task_name(finished_task, "provider-connection-check")
response = authenticated_client_no_permissions_rbac.delete(
reverse("task-detail", kwargs={"pk": finished_task.id})
)
assert response.status_code == status.HTTP_403_FORBIDDEN
@patch("api.v1.views.AsyncResult")
def test_tasks_revoke_unauthenticated_returns_401(
self, mock_async_result, tasks_fixture
):
from rest_framework.test import APIClient
_, pending_task = tasks_fixture
self._set_task_name(pending_task, "scan-perform")
response = APIClient().delete(
reverse("task-detail", kwargs={"pk": pending_task.id})
)
assert response.status_code == status.HTTP_401_UNAUTHORIZED
mock_async_result.return_value.revoke.assert_not_called()
@patch("api.v1.views.AsyncResult")
def test_tasks_revoke_foreign_tenant_task_returns_404(
self,
mock_async_result,
authenticated_client_for_tenant_factory,
tenants_fixture,
tasks_fixture,
):
_, foreign_tenant, *_ = tenants_fixture
_, pending_task = tasks_fixture
self._set_task_name(pending_task, "scan-perform")
client = self._client_with_role(
foreign_tenant, authenticated_client_for_tenant_factory, manage_scans=True
)
response = client.delete(reverse("task-detail", kwargs={"pk": pending_task.id}))
assert response.status_code == status.HTTP_404_NOT_FOUND
mock_async_result.return_value.revoke.assert_not_called()
@pytest.mark.django_db
class TestAttackPathsScanViewSet:
+7 -14
View File
@@ -302,26 +302,17 @@ def get_prowler_provider_kwargs(
def _normalize_oraclecloud_provider_kwargs(secret: dict) -> dict:
"""Normalize external OCI secret fields into SDK provider kwargs."""
prowler_provider_kwargs = secret.copy()
home_region = _oraclecloud_home_region(prowler_provider_kwargs.pop("region", None))
if home_region:
prowler_provider_kwargs["home_region"] = home_region
prowler_provider_kwargs.pop("region", None)
return prowler_provider_kwargs
def _oraclecloud_home_region(region) -> str | None:
"""Return the stored OCI region as a home region, ignoring blank or non-string legacy values."""
if isinstance(region, str) and region.strip():
return region.strip()
return None
def _normalize_oraclecloud_connection_test_kwargs(secret: dict) -> dict:
"""Normalize external OCI secret fields into test_connection kwargs."""
from prowler.providers.oraclecloud.oraclecloud_provider import OraclecloudProvider
prowler_provider_kwargs = secret.copy()
home_region = _oraclecloud_home_region(prowler_provider_kwargs.pop("region", None))
prowler_provider_kwargs.pop("region", None)
if (
prowler_provider_kwargs.get("user")
@@ -332,9 +323,11 @@ def _normalize_oraclecloud_connection_test_kwargs(secret: dict) -> dict:
or prowler_provider_kwargs.get("key_file")
)
):
# Identity calls only succeed in a region the tenancy is subscribed to.
prowler_provider_kwargs["region"] = (
home_region or OraclecloudProvider._bootstrap_region
# Connection validation needs one OCI endpoint, but scans remain unfiltered.
prowler_provider_kwargs["region"] = getattr(
OraclecloudProvider,
"_bootstrap_region",
OraclecloudProvider._home_region,
)
return prowler_provider_kwargs
@@ -301,7 +301,8 @@ from rest_framework_json_api import serializers
},
"region": {
"type": "string",
"description": "Optional OCI home region (or any region the tenancy is subscribed to) used to validate the credentials. It does not filter the scan, which covers all subscribed regions. Defaults to us-ashburn-1.",
"deprecated": True,
"description": "Legacy OCI region field accepted for backwards compatibility but ignored; OCI scans all regions.",
},
},
"required": ["user", "fingerprint", "tenancy"],
+4 -13
View File
@@ -71,7 +71,6 @@ from django.db import IntegrityError, transaction
from drf_spectacular.utils import extend_schema_field
from jwt.exceptions import InvalidKeyError
from prowler.lib.mutelist.mutelist import Mutelist
from prowler.providers.oraclecloud.config import OCI_REGIONS
from rest_framework.reverse import reverse
from rest_framework.validators import UniqueTogetherValidator
from rest_framework_json_api import serializers
@@ -1918,16 +1917,9 @@ class IacProviderSecret(serializers.Serializer):
resource_name = "provider-secrets"
class OCIHomeRegionField(serializers.Field):
"""Optional OCI home region; blank or non-string legacy values are dropped."""
class LegacyOCIRegionField(serializers.Field):
def to_internal_value(self, data):
if not isinstance(data, str) or not data.strip():
return None
region = data.strip()
if region not in OCI_REGIONS:
raise serializers.ValidationError(f"Invalid OCI region: {region}")
return region
return data
def to_representation(self, value):
return value
@@ -1940,11 +1932,10 @@ class OracleCloudProviderSecret(serializers.Serializer):
key_content = serializers.CharField(required=False)
tenancy = serializers.CharField()
pass_phrase = serializers.CharField(required=False)
region = OCIHomeRegionField(required=False, allow_null=True)
region = LegacyOCIRegionField(required=False, allow_null=True)
def validate(self, attrs):
if not attrs.get("region"):
attrs.pop("region", None)
attrs.pop("region", None)
if "key_file" not in attrs and "key_content" not in attrs:
raise serializers.ValidationError(
+81 -81
View File
@@ -125,14 +125,10 @@ from api.models import (
)
from api.pagination import ComplianceOverviewPagination
from api.rbac.permissions import (
TASK_REVOKE_PERMISSIONS,
Permissions,
get_integrations,
get_providers,
get_role,
get_tasks,
get_user_roles,
roles_have_permissions,
)
from api.renderers import APIJSONRenderer, PlainTextRenderer
from api.rls import Tenant
@@ -2826,15 +2822,6 @@ class ScanViewSet(ProviderVisibilityMixin, BaseRLSViewSet):
tenant_id=self.request.tenant_id,
task_id=pre_task_id,
task_status=(QUEUED_SCAN_TASK_STATE if active_scan else None),
# This response is serialized before the on_commit publish,
# so without these the caller gets a task id and no scan id.
# Kept in step with what `enqueue_scan_execution_on_commit`
# publishes below.
task_kwargs={
"tenant_id": str(self.request.tenant_id),
"scan_id": str(scan.id),
"provider_id": str(scan.provider_id),
},
)
if not active_scan:
@@ -2862,29 +2849,17 @@ class ScanViewSet(ProviderVisibilityMixin, BaseRLSViewSet):
list=extend_schema(
tags=["Task"],
summary="List all tasks",
description=(
"Retrieve a list of all tasks with options for filtering by name, state, and other "
"criteria. Tasks that reference a provider are only returned when the role can "
"access it; tasks without a provider reference are returned for every role."
),
description="Retrieve a list of all tasks with options for filtering by name, state, and other criteria.",
),
retrieve=extend_schema(
tags=["Task"],
summary="Retrieve data from a specific task",
description=(
"Fetch detailed information about a specific task by its ID. Tasks tied to a provider "
"outside the visibility of the role are not found."
),
description="Fetch detailed information about a specific task by its ID.",
),
destroy=extend_schema(
tags=["Task"],
summary="Revoke a task",
description=(
"Try to revoke a task using its ID. Only tasks that are not yet in progress can be "
"revoked, and the caller needs the same permission as the operation that queued "
"the task (for example MANAGE_SCANS for a scan). Provider deletions cannot be "
"revoked."
),
description="Try to revoke a task using its ID. Only tasks that are not yet in progress can be revoked.",
responses={202: OpenApiResponse(response=TaskSerializer)},
),
)
@@ -2900,26 +2875,13 @@ class TaskViewSet(BaseRLSViewSet):
required_permissions = []
def get_queryset(self):
return (
get_tasks(self.user_role)
.annotate(
name=F("task_runner_task__task_name"),
state=F("task_runner_task__status"),
)
.select_related("task_runner_task")
)
return Task.objects.annotate(
name=F("task_runner_task__task_name"),
state=F("task_runner_task__status"),
).select_related("task_runner_task")
def destroy(self, request, *args, pk=None, **kwargs):
task = self.get_object()
required_permissions = TASK_REVOKE_PERMISSIONS.get(
task.task_runner_task.task_name
)
# Same multi-role semantics as HasPermissions.
if required_permissions is None or not roles_have_permissions(
get_user_roles(request.user, request.tenant_id), required_permissions
):
raise PermissionDenied("You do not have permission to revoke this task.")
task = get_object_or_404(Task, pk=pk)
if task.task_runner_task.status not in ["PENDING", "RECEIVED"]:
serializer = TaskSerializer(task)
return Response(
@@ -3516,8 +3478,12 @@ class ResourceViewSet(PaginateByPkMixin, BaseRLSViewSet):
filtered_queryset = self.filter_queryset(self.get_queryset())
latest_scans = (
Scan.all_objects.filter(tenant_id=tenant_id)
.latest_per_provider()
Scan.all_objects.filter(
tenant_id=tenant_id,
state=StateChoices.COMPLETED,
)
.order_by("provider_id", "-inserted_at")
.distinct("provider_id")
.values("provider_id")
)
@@ -3649,9 +3615,11 @@ class ResourceViewSet(PaginateByPkMixin, BaseRLSViewSet):
tenant_id = request.tenant_id
query_params = request.query_params
latest_scans_queryset = Scan.all_objects.filter(
tenant_id=tenant_id
).latest_per_provider()
latest_scans_queryset = (
Scan.all_objects.filter(tenant_id=tenant_id, state=StateChoices.COMPLETED)
.order_by("provider_id", "-inserted_at")
.distinct("provider_id")
)
queryset = ResourceScanSummary.objects.filter(
tenant_id=tenant_id,
@@ -4238,9 +4206,12 @@ class FindingViewSet(PaginateByPkMixin, BaseRLSViewSet):
tenant_id = request.tenant_id
filtered_queryset = self.filter_queryset(self.get_queryset())
latest_scan_ids = Scan.all_objects.filter(
tenant_id=tenant_id
).latest_ids_per_provider()
latest_scan_ids = list(
Scan.all_objects.filter(tenant_id=tenant_id, state=StateChoices.COMPLETED)
.order_by("provider_id", "-inserted_at")
.distinct("provider_id")
.values_list("id", flat=True)
)
filtered_queryset = filtered_queryset.filter(
tenant_id=tenant_id, scan_id__in=latest_scan_ids
)
@@ -4263,9 +4234,11 @@ class FindingViewSet(PaginateByPkMixin, BaseRLSViewSet):
tenant_id = request.tenant_id
query_params = request.query_params
latest_scans_queryset = Scan.all_objects.filter(
tenant_id=tenant_id
).latest_per_provider()
latest_scans_queryset = (
Scan.all_objects.filter(tenant_id=tenant_id, state=StateChoices.COMPLETED)
.order_by("provider_id", "-inserted_at")
.distinct("provider_id")
)
raw_latest_scans_ids = list(
latest_scans_queryset.values_list("id", "unique_resource_count")
)
@@ -5006,7 +4979,11 @@ class ComplianceOverviewViewSet(
if provider_filters:
scans = scans.filter(**provider_filters)
return scans.latest_ids_per_provider()
return list(
scans.order_by("provider_id", "-inserted_at")
.distinct("provider_id")
.values_list("id", flat=True)
)
def _filtered_queryset_for_latest_provider_scans(self, latest_scan_ids=None):
if latest_scan_ids is None:
@@ -5748,9 +5725,14 @@ class OverviewViewSet(ProviderFilterParamsMixin, BaseRLSViewSet):
else {}
)
latest_scan_ids = Scan.all_objects.filter(
tenant_id=tenant_id, **provider_filter
).latest_ids_per_provider()
latest_scan_ids = (
Scan.all_objects.filter(
tenant_id=tenant_id, state=StateChoices.COMPLETED, **provider_filter
)
.order_by("provider_id", "-inserted_at")
.distinct("provider_id")
.values_list("id", flat=True)
)
return filtered_queryset.filter(
tenant_id=tenant_id, scan_id__in=latest_scan_ids
@@ -5777,10 +5759,16 @@ class OverviewViewSet(ProviderFilterParamsMixin, BaseRLSViewSet):
def _latest_scan_ids_for_allowed_providers(self, tenant_id, provider_filters=None):
provider_filter = self._get_provider_filter()
queryset = Scan.all_objects.filter(tenant_id=tenant_id, **provider_filter)
queryset = Scan.all_objects.filter(
tenant_id=tenant_id, state=StateChoices.COMPLETED, **provider_filter
)
if provider_filters:
queryset = queryset.filter(**provider_filters)
return queryset.latest_ids_per_provider()
return (
queryset.order_by("provider_id", "-inserted_at")
.distinct("provider_id")
.values_list("id", flat=True)
)
@action(detail=False, methods=["get"], url_name="providers")
def providers(self, request):
@@ -5792,9 +5780,14 @@ class OverviewViewSet(ProviderFilterParamsMixin, BaseRLSViewSet):
else {}
)
latest_scan_ids = Scan.all_objects.filter(
tenant_id=tenant_id, **provider_filter
).latest_ids_per_provider()
latest_scan_ids = (
Scan.all_objects.filter(
tenant_id=tenant_id, state=StateChoices.COMPLETED, **provider_filter
)
.order_by("provider_id", "-inserted_at")
.distinct("provider_id")
.values_list("id", flat=True)
)
findings_aggregated = (
queryset.filter(scan_id__in=latest_scan_ids)
@@ -7941,13 +7934,18 @@ class FindingGroupViewSet(JsonApiFilterMixin, BaseRLSViewSet):
def _get_latest_findings_per_provider(self, filtered_queryset):
"""Keep only findings from each provider's most recent completed scan."""
# `latest_ids_per_provider` materializes to a literal IN list.
# Left as a subquery, Postgres can't estimate the match count and picks
# a serial nested loop on resource_finding_mappings when one scan
# dominates findings
latest_scan_ids = Scan.objects.filter(
tenant_id=self.request.tenant_id
).latest_ids_per_provider()
# Materialize to a literal IN list. Left as a subquery, Postgres can't
# estimate the match count and picks a serial nested loop on
# resource_finding_mappings when one scan dominates findings
latest_scan_ids = list(
Scan.objects.filter(
tenant_id=self.request.tenant_id,
state=StateChoices.COMPLETED,
)
.order_by("provider_id", "-completed_at", "-inserted_at")
.distinct("provider_id")
.values_list("id", flat=True)
)
return filtered_queryset.filter(scan_id__in=latest_scan_ids)
def _post_process_aggregation(self, aggregated_data):
@@ -8893,14 +8891,16 @@ class FindingGroupViewSet(JsonApiFilterMixin, BaseRLSViewSet):
tenant_id = request.tenant_id
queryset = self._get_finding_queryset()
# The shared selector orders by -completed_at (matching the /latest
# summary path and the daily summary upsert keyed on
# midnight(completed_at)) so that overlapping scans do not make
# /resources and /latest read from different scans and report
# diverging counts.
latest_scan_ids = Scan.objects.filter(
tenant_id=tenant_id
).latest_ids_per_provider()
# Order by -completed_at (matching the /latest summary path and the
# daily summary upsert keyed on midnight(completed_at)) so that
# overlapping scans do not make /resources and /latest read from
# different scans and report diverging counts.
latest_scan_ids = (
Scan.objects.filter(tenant_id=tenant_id, state=StateChoices.COMPLETED)
.order_by("provider_id", "-completed_at", "-inserted_at")
.distinct("provider_id")
.values_list("id", flat=True)
)
normalized_params = self._normalize_jsonapi_params(request.query_params)
# Remove date filters since we're using latest
-56
View File
@@ -231,62 +231,6 @@ LOGGING = {
"level": LEVEL,
"propagate": False,
},
# Celery loggers must be declared explicitly because
# disable_existing_loggers=True silences any logger that exists at
# dictConfig time but is not named here. Without these, fatal worker
# errors (e.g. celery.worker CRITICAL) produce no output.
# "celery" must keep propagating: get_task_logger() parents task
# loggers under celery.task, so blocking here hides them from root.
"celery": {
"level": LEVEL,
"propagate": True,
},
"celery.worker": {
"handlers": ["tasks_console"],
"level": LEVEL,
"propagate": False,
},
"celery.worker.consumer": {
"handlers": ["tasks_console"],
"level": LEVEL,
"propagate": False,
},
"celery.worker.consumer.consumer": {
"handlers": ["tasks_console"],
"level": LEVEL,
"propagate": False,
},
"kombu": {
"handlers": ["tasks_console"],
"level": LEVEL,
"propagate": False,
},
"kombu.transport.redis": {
"handlers": ["tasks_console"],
"level": LEVEL,
"propagate": False,
},
"billiard": {
"handlers": ["tasks_console"],
"level": LEVEL,
"propagate": False,
},
"amqp": {
"handlers": ["tasks_console"],
"level": LEVEL,
"propagate": False,
},
# WARNING keeps task failures but skips one "succeeded" line per task.
"celery.app.trace": {
"handlers": ["tasks_console"],
"level": "WARNING",
"propagate": False,
},
"celery.beat": {
"handlers": ["tasks_console"],
"level": LEVEL,
"propagate": False,
},
},
# Gunicorn required configuration
"root": {
-15
View File
@@ -7,7 +7,6 @@ from config.settings.eventstream import * # noqa
from config.settings.partitions import * # noqa
from config.settings.sentry import * # noqa
from config.settings.social_login import * # noqa
from django.core.exceptions import ImproperlyConfigured
SECRET_KEY = env("SECRET_KEY", default="secret")
DEBUG = env.bool("DJANGO_DEBUG", default=False)
@@ -296,9 +295,6 @@ DJANGO_OUTPUT_S3_AWS_SECRET_ACCESS_KEY = env.str(
)
DJANGO_OUTPUT_S3_AWS_SESSION_TOKEN = env.str("DJANGO_OUTPUT_S3_AWS_SESSION_TOKEN", "")
DJANGO_OUTPUT_S3_AWS_DEFAULT_REGION = env.str("DJANGO_OUTPUT_S3_AWS_DEFAULT_REGION", "")
# Storage endpoint the API and Celery workers use to talk to S3-compatible object storage
# such as MinIO. Empty means the real AWS S3 endpoint, which is unaffected.
DJANGO_OUTPUT_S3_AWS_ENDPOINT_URL = env.str("DJANGO_OUTPUT_S3_AWS_ENDPOINT_URL", "")
# Browser-reachable storage host used to sign download URLs. Empty means sign against the
# same endpoint the API talks to, which is what Prowler Cloud on S3 does.
DJANGO_OUTPUT_S3_AWS_PUBLIC_ENDPOINT_URL = env.str(
@@ -329,17 +325,6 @@ ATTACK_PATHS_SCAN_STALE_THRESHOLD_MINUTES = env.int(
"ATTACK_PATHS_SCAN_STALE_THRESHOLD_MINUTES", 960
) # 16h
# Minimum age (of the scan row, or of the scan id itself when the row is gone) before
# the periodic reaper will drop an orphaned temp Neo4j database. Keeps a scan that is
# still legitimately in flight from ever losing its staging database mid-run.
ATTACK_PATHS_TMP_DB_REAP_SAFETY_MARGIN_HOURS = env.int(
"ATTACK_PATHS_TMP_DB_REAP_SAFETY_MARGIN_HOURS", 6
)
if ATTACK_PATHS_TMP_DB_REAP_SAFETY_MARGIN_HOURS <= 0:
raise ImproperlyConfigured(
"ATTACK_PATHS_TMP_DB_REAP_SAFETY_MARGIN_HOURS must be a positive number of hours"
)
# Selects where the persistent attack-paths graph is stored. The scan
# temporary database is always Neo4j; only the sink is configurable.
# Valid values: "neo4j" (default, OSS and local dev), "neptune" (hosted).
@@ -1,107 +0,0 @@
"""Periodic reaper for orphaned temp Neo4j scan databases.
`scan.py` creates a throw-away `db-tmp-scan-<attack_paths_scan_id>` database per
scan and drops it once the scan finishes, success or failure. When the worker
or Neo4j itself dies mid-scan, that drop never runs and nothing else ever
revisits the database - it sits there forever. This sweep lists every temp
database on the ingest cluster and drops the ones whose scan is gone or has
been finished for longer than the configured safety margin.
"""
from datetime import UTC, datetime, timedelta
from api.attack_paths import database as graph_database
from api.db_router import MainRouter
from api.models import AttackPathsScan, StateChoices
from api.uuid_utils import datetime_from_uuid7
from celery.utils.log import get_task_logger
from config.django.base import ATTACK_PATHS_TMP_DB_REAP_SAFETY_MARGIN_HOURS
from uuid6 import UUID as UUID7
logger = get_task_logger(__name__)
TERMINAL_STATES = (
StateChoices.COMPLETED,
StateChoices.FAILED,
StateChoices.CANCELLED,
)
def reap_orphaned_tmp_databases() -> dict:
"""Drop temp Neo4j scan databases whose scan is gone or long finished.
A failure listing databases aborts the whole sweep (nothing to iterate).
A failure reaping one database is logged and skipped so the rest of the
sweep still runs.
"""
now = datetime.now(tz=UTC)
safety_margin = timedelta(hours=ATTACK_PATHS_TMP_DB_REAP_SAFETY_MARGIN_HOURS)
try:
databases = graph_database.list_databases()
except Exception:
logger.exception("Failed to list ingest Neo4j databases for temp-db reap")
return {"dropped_count": 0, "databases": []}
tmp_databases = [
name for name in databases if name.startswith(graph_database.TEMP_DB_PREFIX)
]
dropped: list[str] = []
for database in tmp_databases:
try:
if _is_orphaned(database, now, safety_margin):
graph_database.drop_database(database)
dropped.append(database)
logger.info(f"Dropped orphaned temp Neo4j database `{database}`")
except Exception:
logger.exception(f"Failed to reap temp Neo4j database `{database}`")
logger.info(f"Temp Neo4j database reap: {len(dropped)} dropped")
return {"dropped_count": len(dropped), "databases": dropped}
def _is_orphaned(database: str, now: datetime, safety_margin: timedelta) -> bool:
"""Decide whether a temp database is safe to drop.
No scan row: the row was hard-deleted (tenant/provider cleanup) or was
never created. Falls back to the scan id's own UUIDv7 timestamp so a
database created moments ago is never touched even without a row to check.
Scan row present: only reapable once it reached a terminal state and has
been finished for longer than the safety margin, so a scan still
legitimately executing is never touched.
"""
scan_id = database[len(graph_database.TEMP_DB_PREFIX) :]
try:
scan_uuid = UUID7(scan_id)
except ValueError:
logger.warning(
f"Temp database `{database}` has an unparseable scan id, skipping"
)
return False
# Global sweep with no tenant context: admin_db bypasses RLS on purpose, the same
# way cleanup_stale_attack_paths_scans finds stale scans across every tenant.
scan = (
AttackPathsScan.all_objects.using(MainRouter.admin_db)
.filter(id=scan_uuid)
.first()
)
if scan is None:
if scan_uuid.version != 7:
logger.warning(
f"Temp database `{database}` has no scan row and a non-UUIDv7 id, "
"skipping"
)
return False
return now - datetime_from_uuid7(scan_uuid) >= safety_margin
if scan.state not in TERMINAL_STATES:
return False
# `mark_scan_finished` does not touch `updated_at`, so prefer `completed_at`
finished_at = scan.completed_at or scan.updated_at
return now - finished_at >= safety_margin
+3 -6
View File
@@ -499,13 +499,10 @@ def backfill_provider_compliance_scores(tenant_id: str) -> dict:
provider_id__in=existing_providers
)
# `completed_scans` keeps its own `completed_at__isnull=False`: this
# task writes a *dated* ProviderComplianceScore row, so unlike the read
# paths it genuinely cannot use a scan without a `completed_at`.
scan_info = list(
completed_scans.latest_per_provider().values(
"id", "provider_id", "completed_at"
)
completed_scans.order_by("provider_id", "-completed_at")
.distinct("provider_id")
.values("id", "provider_id", "completed_at")
)
if not scan_info:
+3 -14
View File
@@ -13,7 +13,6 @@ from api.models import (
Tenant,
)
from celery.utils.log import get_task_logger
from django.conf import settings
from django.db import DatabaseError
from tasks.jobs.queries import (
COMPLIANCE_DELETE_EMPTY_TENANT_SUMMARY_SQL,
@@ -107,19 +106,9 @@ def delete_provider(tenant_id: str, pk: str):
try:
if attack_paths_sink_backends:
for sink_backend in attack_paths_sink_backends:
try:
backend = sink_module.get_backend_for_name(sink_backend)
except RuntimeError as sink_error:
# A retired sink has no connection settings left, and no graph left to drop
if sink_backend == settings.ATTACK_PATHS_SINK_DATABASE.lower():
raise
logger.warning(
f"Skipping graph cleanup on unconfigured sink {sink_backend}: {sink_error}"
)
continue
backend.drop_subgraph(tenant_database_name, str(pk))
sink_module.get_backend_for_name(sink_backend).drop_subgraph(
tenant_database_name, str(pk)
)
else:
graph_database.drop_subgraph(tenant_database_name, str(pk))
+10 -28
View File
@@ -207,21 +207,15 @@ def get_s3_client():
This function attempts to initialize an S3 client by reading the AWS access key, secret key,
session token, and region from environment variables. It then validates the client by listing
available S3 buckets. If an error occurs during this process (for example, due to missing or
invalid credentials), it falls back to creating an S3 client without explicitly provided
credentials, which may rely on other configuration sources (e.g., IAM roles).
That fallback is only safe when no explicit endpoint is configured: with an endpoint set, the
explicit client already targets the intended S3-compatible storage, and the fallback client
would go to the AWS default provider chain instead, an unrelated real-AWS account reachable
from the host. So when an endpoint is configured, the original error propagates instead.
invalid credentials), it falls back to creating an S3 client without explicitly provided credentials,
which may rely on other configuration sources (e.g., IAM roles).
Returns:
boto3.client: A configured S3 client instance.
Raises:
ClientError, NoCredentialsError, or ParamValidationError if the client cannot be created.
ClientError, NoCredentialsError, or ParamValidationError if both attempts to create a client fail.
"""
endpoint = settings.DJANGO_OUTPUT_S3_AWS_ENDPOINT_URL
s3_client = None
try:
s3_client = boto3.client(
@@ -232,12 +226,9 @@ def get_s3_client():
# Storage that has no meaningful region, MinIO among it, is usually configured
# without one, and botocore rejects an empty region before any request is made.
region_name=settings.DJANGO_OUTPUT_S3_AWS_DEFAULT_REGION or "us-east-1",
endpoint_url=endpoint or None,
)
s3_client.list_buckets()
except (ClientError, NoCredentialsError, ParamValidationError, ValueError):
if endpoint:
raise
s3_client = boto3.client("s3")
s3_client.list_buckets()
@@ -245,21 +236,13 @@ def get_s3_client():
def get_s3_presign_client():
"""Return a client that signs download URLs with SigV4.
"""Return a client that signs URLs against the public storage host.
It is used when a public or internal storage host is configured, or when the bucket's
region is: boto3 otherwise presigns S3 URLs with SigV2, which S3 rejects for SSE-KMS
objects. None means none of those is set and the caller should presign with its own
client, which leaves those deployments with the URL they get today.
The public endpoint wins when both are set: the internal endpoint may only be reachable
from inside the cluster, and a URL signed against it would not open in a browser.
None means no public host is configured and the caller should presign with its own
client, which leaves deployments on real S3 with the URL they get today.
"""
endpoint = (
settings.DJANGO_OUTPUT_S3_AWS_PUBLIC_ENDPOINT_URL
or settings.DJANGO_OUTPUT_S3_AWS_ENDPOINT_URL
)
if not endpoint and not settings.DJANGO_OUTPUT_S3_AWS_DEFAULT_REGION:
public_endpoint = settings.DJANGO_OUTPUT_S3_AWS_PUBLIC_ENDPOINT_URL
if not public_endpoint:
return None
# Blank keys are signed as-is (empty credential scope) instead of deferring to the
@@ -283,10 +266,9 @@ def get_s3_presign_client():
# SigV4 puts the region in the credential scope, and MinIO answers to us-east-1
# unless it was told otherwise, so an empty region would sign an unusable URL.
region_name=settings.DJANGO_OUTPUT_S3_AWS_DEFAULT_REGION or "us-east-1",
endpoint_url=endpoint or None,
endpoint_url=public_endpoint,
# The signature covers the host, so the addressing style has to be pinned rather
# than guessed from the endpoint: MinIO serves path-style, and on AWS it keeps the
# regional host instead of the global one, which redirects for new buckets.
# than guessed from the endpoint: MinIO serves path-style.
config=Config(signature_version="s3v4", s3={"addressing_style": "path"}),
)
+12 -4
View File
@@ -1,7 +1,7 @@
from collections.abc import Iterable
from api.db_utils import rls_transaction
from api.models import Finding, MuteRule, Scan
from api.models import Finding, MuteRule, Scan, StateChoices
from celery.utils.log import get_task_logger
logger = get_task_logger(__name__)
@@ -39,9 +39,17 @@ def mute_findings_in_latest_scans(
with rls_transaction(tenant_id):
mute_rule = MuteRule.objects.get(id=mute_rule_id, tenant_id=tenant_id)
latest_scans = Scan.objects.filter(
tenant_id=tenant_id, provider_id__in=provider_ids
).latest_ids_per_provider()
latest_scans = list(
Scan.objects.filter(
tenant_id=tenant_id,
provider_id__in=provider_ids,
state=StateChoices.COMPLETED,
completed_at__isnull=False,
)
.order_by("provider_id", "-completed_at", "-inserted_at", "-id")
.distinct("provider_id")
.values_list("id", flat=True)
)
changed_scan_ids = []
findings_muted = 0
@@ -84,7 +84,6 @@ _SKIP_RECOVERY = {
"scan-perform-scheduled",
"attack-paths-scan-perform",
"attack-paths-cleanup-stale-scans",
"attack-paths-reap-orphaned-tmp-databases",
"reconcile-orphan-tasks",
}
+15 -34
View File
@@ -2113,9 +2113,7 @@ def aggregate_attack_surface(tenant_id: str, scan_id: str):
.annotate(
total=Count("id"),
failed=Count("id", filter=Q(status="FAIL", muted=False)),
# Not `muted`: an annotation named after a model field comes
# back as `muted_new` from the psqlextra queryset.
muted_count=Count("id", filter=Q(status="FAIL", muted=True)),
muted=Count("id", filter=Q(status="FAIL", muted=True)),
)
)
@@ -2126,7 +2124,7 @@ def aggregate_attack_surface(tenant_id: str, scan_id: str):
aggregated_counts[attack_surface_type]["total"] += stats["total"] or 0
aggregated_counts[attack_surface_type]["failed"] += stats["failed"] or 0
aggregated_counts[attack_surface_type]["muted"] += stats["muted_count"] or 0
aggregated_counts[attack_surface_type]["muted"] += stats["muted"] or 0
overview_objects = []
for attack_surface_type, counts in aggregated_counts.items():
@@ -2702,37 +2700,20 @@ def reset_ephemeral_resource_findings_count(tenant_id: str, scan_id: str) -> dic
# refreshed). Wiping based on the older scan would zero counts the newer
# scan just set. Skip and let the newer scan's reset task do the work; if
# this task was delayed in the queue, that's the correct outcome.
#
# The comparison must be against the newest *full-scope* scan, which is
# what this variable has always been named after but did not use to be:
# the query filtered nothing about scope, so any newer scan that is not
# full-scope (an imported one, for instance) made the full-scope scan
# skip its own cleanup and leave ephemeral resources with a stale
# failed_findings_count permanently.
#
# `is_full_scope()` reads `trigger` plus the scoping keys inside
# `scanner_args`, which is not expressible as a WHERE clause, so the
# candidates are walked newest-first in Python until the first full-scope
# one. The walk needs no cap: `scan` is itself a full-scope candidate, so
# it stops at `scan` at the latest, after reading only the scans newer than
# it. A fixed window would return None once more newer scoped scans had
# landed than it inspected, and skip the cleanup exactly like the bug above.
#
# NULL `completed_at` no longer needs an explicit filter here: the shared
# ordering in `ScanQuerySet.LATEST_ORDER_BY` sorts NULLs last
# rather than excluding them, which also fixes the case where a provider
# whose completed scans all have a NULL `completed_at` resolved to None and
# therefore never ran the reset at all.
# `completed_at__isnull=False` is required: Postgres orders NULL first in
# DESC, so a sibling COMPLETED scan with a missing completed_at would sort
# as "newest" and incorrectly cause us to skip.
with rls_transaction(tenant_id):
candidates = (
Scan.objects.filter(tenant_id=tenant_id, provider_id=scan.provider_id)
.latest_first()
.only("id", "trigger", "scanner_args")
.iterator(chunk_size=100)
)
latest_full_scope_scan_id = next(
(candidate.id for candidate in candidates if candidate.is_full_scope()),
None,
latest_full_scope_scan_id = (
Scan.objects.filter(
tenant_id=tenant_id,
provider_id=scan.provider_id,
state=StateChoices.COMPLETED,
completed_at__isnull=False,
)
.order_by("-completed_at", "-inserted_at")
.values_list("id", flat=True)
.first()
)
if latest_full_scope_scan_id != scan.id:
logger.info(
+1 -25
View File
@@ -1,4 +1,3 @@
import json
import os
from datetime import UTC, datetime, timedelta
from pathlib import Path
@@ -43,7 +42,6 @@ from tasks.jobs.attack_paths import (
)
from tasks.jobs.attack_paths import db_utils as attack_paths_db_utils
from tasks.jobs.attack_paths.cleanup import cleanup_stale_attack_paths_scans
from tasks.jobs.attack_paths.tmp_db_reaper import reap_orphaned_tmp_databases
from tasks.jobs.backfill import (
aggregate_scan_category_summaries,
aggregate_scan_resource_group_summaries,
@@ -165,28 +163,13 @@ def create_scan_task_record(
task_id: str,
task_name: str = "scan-perform",
task_status: str | None = states.PENDING,
task_kwargs: dict | None = None,
) -> Task:
"""Pre-create the TaskResult + Task rows for a pre-generated task id.
Pass ``task_kwargs`` when the response built from this record is serialized
before the broker publish. ``task_kwargs`` is otherwise only written by the
``before_task_publish`` signal (``api/signals.py``), and the scan publish is
deferred to ``on_commit``, so the 202 would carry an empty ``task_args`` and
the caller would have no way to learn the scan id it was just handed a task
for. The publish later overwrites the field with the same kwargs as a Python
repr; both forms decode to the same dict (``decode_celery_field``).
"""
if task_status is None:
task_status = states.PENDING
defaults = {"status": task_status, "task_name": task_name}
if task_kwargs is not None:
defaults["task_kwargs"] = json.dumps(task_kwargs)
task_result, _ = TaskResult.objects.update_or_create(
task_id=str(task_id),
defaults=defaults,
defaults={"status": task_status, "task_name": task_name},
)
prowler_task, _ = Task.objects.update_or_create(
id=str(task_id),
@@ -728,13 +711,6 @@ def cleanup_stale_attack_paths_scans_task():
return cleanup_stale_attack_paths_scans()
@shared_task(
name="attack-paths-reap-orphaned-tmp-databases", queue="attack-paths-scans"
)
def reap_orphaned_attack_paths_tmp_databases_task():
return reap_orphaned_tmp_databases()
@shared_task(name="reconcile-orphan-tasks", queue="celery")
def reconcile_orphan_tasks_task():
"""Periodic watchdog: recover tasks whose worker is gone (deploys, crashes)."""
@@ -1,286 +0,0 @@
from datetime import UTC, datetime, timedelta
from unittest.mock import patch
from uuid import uuid4
import pytest
from api.attack_paths.database import TEMP_DB_PREFIX
from api.models import AttackPathsScan, StateChoices
from api.uuid_utils import datetime_to_uuid7
from config.django.base import ATTACK_PATHS_TMP_DB_REAP_SAFETY_MARGIN_HOURS
MARGIN = timedelta(hours=ATTACK_PATHS_TMP_DB_REAP_SAFETY_MARGIN_HOURS)
def _tmp_db_name(scan_uuid) -> str:
return f"{TEMP_DB_PREFIX}{scan_uuid}"
@pytest.mark.django_db
class TestReapOrphanedTmpDatabases:
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.drop_database")
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.list_databases")
def test_ignores_databases_without_the_temp_prefix(self, mock_list, mock_drop):
from tasks.jobs.attack_paths.tmp_db_reaper import reap_orphaned_tmp_databases
mock_list.return_value = ["db-tenant-abc123", "system", "neo4j"]
result = reap_orphaned_tmp_databases()
assert result == {"dropped_count": 0, "databases": []}
mock_drop.assert_not_called()
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.drop_database")
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.list_databases")
def test_drops_temp_db_with_no_scan_row_past_safety_margin(
self, mock_list, mock_drop
):
from tasks.jobs.attack_paths.tmp_db_reaper import reap_orphaned_tmp_databases
old_scan_id = datetime_to_uuid7(
datetime.now(tz=UTC) - MARGIN - timedelta(hours=1)
)
database = _tmp_db_name(old_scan_id)
mock_list.return_value = [database]
result = reap_orphaned_tmp_databases()
assert result == {"dropped_count": 1, "databases": [database]}
mock_drop.assert_called_once_with(database)
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.drop_database")
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.list_databases")
def test_preserves_temp_db_with_no_scan_row_inside_safety_margin(
self, mock_list, mock_drop
):
from tasks.jobs.attack_paths.tmp_db_reaper import reap_orphaned_tmp_databases
recent_scan_id = datetime_to_uuid7(datetime.now(tz=UTC) - timedelta(minutes=5))
database = _tmp_db_name(recent_scan_id)
mock_list.return_value = [database]
result = reap_orphaned_tmp_databases()
assert result == {"dropped_count": 0, "databases": []}
mock_drop.assert_not_called()
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.drop_database")
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.list_databases")
def test_preserves_temp_db_with_unparseable_scan_id(self, mock_list, mock_drop):
from tasks.jobs.attack_paths.tmp_db_reaper import reap_orphaned_tmp_databases
database = f"{TEMP_DB_PREFIX}not-a-uuid"
mock_list.return_value = [database]
result = reap_orphaned_tmp_databases()
assert result == {"dropped_count": 0, "databases": []}
mock_drop.assert_not_called()
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.drop_database")
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.list_databases")
def test_drops_terminal_scan_past_safety_margin(
self, mock_list, mock_drop, tenants_fixture, aws_provider
):
from tasks.jobs.attack_paths.tmp_db_reaper import reap_orphaned_tmp_databases
tenant = tenants_fixture[0]
old_updated_at = datetime.now(tz=UTC) - MARGIN - timedelta(hours=1)
scan = AttackPathsScan.objects.create(
tenant_id=tenant.id,
provider=aws_provider,
state=StateChoices.COMPLETED,
)
AttackPathsScan.objects.filter(id=scan.id).update(updated_at=old_updated_at)
database = _tmp_db_name(scan.id)
mock_list.return_value = [database]
result = reap_orphaned_tmp_databases()
assert result == {"dropped_count": 1, "databases": [database]}
mock_drop.assert_called_once_with(database)
@pytest.mark.parametrize(
"state",
[StateChoices.FAILED, StateChoices.CANCELLED],
)
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.drop_database")
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.list_databases")
def test_drops_other_terminal_states_past_safety_margin(
self, mock_list, mock_drop, tenants_fixture, aws_provider, state
):
from tasks.jobs.attack_paths.tmp_db_reaper import reap_orphaned_tmp_databases
tenant = tenants_fixture[0]
old_updated_at = datetime.now(tz=UTC) - MARGIN - timedelta(hours=1)
scan = AttackPathsScan.objects.create(
tenant_id=tenant.id,
provider=aws_provider,
state=state,
)
AttackPathsScan.objects.filter(id=scan.id).update(updated_at=old_updated_at)
database = _tmp_db_name(scan.id)
mock_list.return_value = [database]
result = reap_orphaned_tmp_databases()
assert result["dropped_count"] == 1
mock_drop.assert_called_once_with(database)
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.drop_database")
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.list_databases")
def test_preserves_terminal_scan_inside_safety_margin(
self, mock_list, mock_drop, tenants_fixture, aws_provider
):
from tasks.jobs.attack_paths.tmp_db_reaper import reap_orphaned_tmp_databases
tenant = tenants_fixture[0]
scan = AttackPathsScan.objects.create(
tenant_id=tenant.id,
provider=aws_provider,
state=StateChoices.COMPLETED,
)
database = _tmp_db_name(scan.id)
mock_list.return_value = [database]
result = reap_orphaned_tmp_databases()
assert result == {"dropped_count": 0, "databases": []}
mock_drop.assert_not_called()
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.drop_database")
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.list_databases")
def test_margin_counts_from_completed_at_not_updated_at(
self, mock_list, mock_drop, tenants_fixture, aws_provider
):
from tasks.jobs.attack_paths.tmp_db_reaper import reap_orphaned_tmp_databases
tenant = tenants_fixture[0]
old = datetime.now(tz=UTC) - MARGIN - timedelta(hours=1)
scan = AttackPathsScan.objects.create(
tenant_id=tenant.id,
provider=aws_provider,
state=StateChoices.COMPLETED,
)
AttackPathsScan.objects.filter(id=scan.id).update(
updated_at=old, completed_at=datetime.now(tz=UTC)
)
database = _tmp_db_name(scan.id)
mock_list.return_value = [database]
result = reap_orphaned_tmp_databases()
assert result == {"dropped_count": 0, "databases": []}
mock_drop.assert_not_called()
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.drop_database")
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.list_databases")
def test_drops_scan_completed_past_safety_margin(
self, mock_list, mock_drop, tenants_fixture, aws_provider
):
from tasks.jobs.attack_paths.tmp_db_reaper import reap_orphaned_tmp_databases
tenant = tenants_fixture[0]
old = datetime.now(tz=UTC) - MARGIN - timedelta(hours=1)
scan = AttackPathsScan.objects.create(
tenant_id=tenant.id,
provider=aws_provider,
state=StateChoices.FAILED,
)
AttackPathsScan.objects.filter(id=scan.id).update(
updated_at=old, completed_at=old
)
database = _tmp_db_name(scan.id)
mock_list.return_value = [database]
result = reap_orphaned_tmp_databases()
assert result == {"dropped_count": 1, "databases": [database]}
mock_drop.assert_called_once_with(database)
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.drop_database")
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.list_databases")
def test_never_drops_an_executing_scan_regardless_of_age(
self, mock_list, mock_drop, tenants_fixture, aws_provider
):
from tasks.jobs.attack_paths.tmp_db_reaper import reap_orphaned_tmp_databases
tenant = tenants_fixture[0]
very_old = datetime.now(tz=UTC) - timedelta(days=30)
scan = AttackPathsScan.objects.create(
tenant_id=tenant.id,
provider=aws_provider,
state=StateChoices.EXECUTING,
)
AttackPathsScan.objects.filter(id=scan.id).update(updated_at=very_old)
database = _tmp_db_name(scan.id)
mock_list.return_value = [database]
result = reap_orphaned_tmp_databases()
assert result == {"dropped_count": 0, "databases": []}
mock_drop.assert_not_called()
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.drop_database")
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.list_databases")
def test_one_failed_drop_does_not_stop_the_rest_of_the_sweep(
self, mock_list, mock_drop
):
from tasks.jobs.attack_paths.tmp_db_reaper import reap_orphaned_tmp_databases
old_time = datetime.now(tz=UTC) - MARGIN - timedelta(hours=1)
failing_scan_id = datetime_to_uuid7(old_time)
succeeding_scan_id = datetime_to_uuid7(old_time)
failing_db = _tmp_db_name(failing_scan_id)
succeeding_db = _tmp_db_name(succeeding_scan_id)
mock_list.return_value = [failing_db, succeeding_db]
mock_drop.side_effect = [Exception("boom"), None]
result = reap_orphaned_tmp_databases()
assert result == {"dropped_count": 1, "databases": [succeeding_db]}
assert mock_drop.call_count == 2
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.list_databases")
def test_returns_empty_result_when_listing_databases_fails(self, mock_list):
from tasks.jobs.attack_paths.tmp_db_reaper import reap_orphaned_tmp_databases
mock_list.side_effect = Exception("neo4j unreachable")
result = reap_orphaned_tmp_databases()
assert result == {"dropped_count": 0, "databases": []}
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.drop_database")
@patch("tasks.jobs.attack_paths.tmp_db_reaper.graph_database.list_databases")
def test_preserves_temp_db_with_random_uuid_and_no_row(self, mock_list, mock_drop):
"""A non-UUIDv7 id with no matching row has no reliable timestamp, so it
must be left alone rather than guessed at."""
from tasks.jobs.attack_paths.tmp_db_reaper import reap_orphaned_tmp_databases
database = _tmp_db_name(uuid4())
mock_list.return_value = [database]
result = reap_orphaned_tmp_databases()
assert result == {"dropped_count": 0, "databases": []}
mock_drop.assert_not_called()
class TestReapOrphanedTmpDatabasesTask:
@patch(
"tasks.tasks.reap_orphaned_tmp_databases",
return_value={"dropped_count": 2, "databases": ["db-tmp-scan-a"]},
)
def test_task_invokes_the_reaper(self, mock_reap):
from tasks.tasks import reap_orphaned_attack_paths_tmp_databases_task
result = reap_orphaned_attack_paths_tmp_databases_task.run()
assert result == {"dropped_count": 2, "databases": ["db-tmp-scan-a"]}
mock_reap.assert_called_once_with()
+1 -56
View File
@@ -4,7 +4,6 @@ import pytest
from api.attack_paths import database as graph_database
from api.models import Provider, Tenant, TenantComplianceSummary
from django.core.exceptions import ObjectDoesNotExist
from django.test import override_settings
from tasks.jobs.deletion import delete_provider, delete_tenant
@@ -124,60 +123,6 @@ class TestDeleteProvider:
"tenant-db", str(instance.id)
)
@override_settings(ATTACK_PATHS_SINK_DATABASE="neo4j")
def test_delete_provider_skips_unconfigured_retired_sink(
self, aws_provider, create_attack_paths_scan
):
instance = aws_provider
tenant_id = str(instance.tenant_id)
create_attack_paths_scan(instance, sink_backend="neo4j")
create_attack_paths_scan(instance, sink_backend="neptune")
neo4j_backend = MagicMock()
def get_backend_for_name(name):
if name == "neptune":
raise RuntimeError("NEPTUNE_WRITER_ENDPOINT and AWS_REGION must be set")
return neo4j_backend
with (
patch(
"tasks.jobs.deletion.graph_database.get_database_name",
return_value="tenant-db",
),
patch(
"tasks.jobs.deletion.sink_module.get_backend_for_name",
side_effect=get_backend_for_name,
),
patch("tasks.jobs.deletion.graph_database.drop_database"),
):
result = delete_provider(tenant_id, instance.id)
assert result
assert not Provider.all_objects.filter(pk=instance.id).exists()
neo4j_backend.drop_subgraph.assert_called_once_with(
"tenant-db", str(instance.id)
)
@override_settings(ATTACK_PATHS_SINK_DATABASE="neo4j")
def test_delete_provider_raises_when_active_sink_unconfigured(
self, aws_provider, create_attack_paths_scan
):
instance = aws_provider
tenant_id = str(instance.tenant_id)
create_attack_paths_scan(instance, sink_backend="neo4j")
with (
patch(
"tasks.jobs.deletion.sink_module.get_backend_for_name",
side_effect=RuntimeError("NEO4J_HOST / NEO4J_PORT must be set"),
),
patch("tasks.jobs.deletion.graph_database.drop_database"),
pytest.raises(RuntimeError),
):
delete_provider(tenant_id, instance.id)
assert Provider.all_objects.filter(pk=instance.id).exists()
def test_delete_provider_continues_when_temp_db_drop_fails(
self, aws_provider, create_attack_paths_scan
):
@@ -204,10 +149,10 @@ class TestDeleteProvider:
assert result
assert not Provider.all_objects.filter(pk=instance.id).exists()
@pytest.mark.usefixtures("provider_compliance_scores_fixture")
def test_delete_provider_recalculates_tenant_compliance_summary(
self,
aws_provider_pair,
provider_compliance_scores_fixture,
):
instance = aws_provider_pair[0]
tenant_id = instance.tenant_id
+3 -95
View File
@@ -3,10 +3,9 @@ import uuid
import zipfile
from datetime import datetime
from pathlib import Path
from unittest.mock import MagicMock, call, patch
from unittest.mock import MagicMock, patch
from urllib.parse import parse_qs, urlparse
import boto3
import pytest
from botocore.exceptions import ClientError
from django.test import override_settings
@@ -64,46 +63,15 @@ class TestOutputs:
assert mock_boto_client.call_args.kwargs["region_name"] == "us-east-1"
@patch("tasks.jobs.export.boto3.client")
@override_settings(DJANGO_OUTPUT_S3_AWS_ENDPOINT_URL="http://minio:9000")
def test_get_s3_client_passes_the_endpoint_when_set(self, mock_boto_client):
get_s3_client()
assert mock_boto_client.call_args.kwargs["endpoint_url"] == "http://minio:9000"
@patch("tasks.jobs.export.boto3.client")
@override_settings(DJANGO_OUTPUT_S3_AWS_ENDPOINT_URL="")
def test_get_s3_client_endpoint_empty_by_default(self, mock_boto_client):
"""Empty keeps today's behavior: no endpoint override, real S3 is used."""
get_s3_client()
assert mock_boto_client.call_args.kwargs["endpoint_url"] is None
@patch("tasks.jobs.export.boto3.client")
@override_settings(DJANGO_OUTPUT_S3_AWS_ENDPOINT_URL="http://minio:9000")
def test_get_s3_client_does_not_fall_back_when_endpoint_set(self, mock_boto_client):
"""A configured endpoint means the explicit client failed talking to it. The fallback
goes to the default provider chain (e.g. an EC2 instance role) against real AWS, so it
must not be used: the original error propagates instead."""
error = ClientError({"Error": {"Code": "403"}}, "ListBuckets")
mock_boto_client.side_effect = error
with pytest.raises(ClientError):
get_s3_client()
mock_boto_client.assert_called_once()
@patch("tasks.jobs.export.boto3.client")
@patch("tasks.jobs.export.settings")
def test_get_s3_client_fallback(self, mock_settings, mock_boto_client):
mock_settings.DJANGO_OUTPUT_S3_AWS_ENDPOINT_URL = ""
mock_boto_client.side_effect = [
ClientError({"Error": {"Code": "403"}}, "ListBuckets"),
MagicMock(),
]
client = get_s3_client()
assert client is not None
assert mock_boto_client.call_args_list[1] == call("s3")
@patch("tasks.jobs.export.get_s3_client")
@patch("tasks.jobs.export.base")
@@ -310,69 +278,9 @@ def _presign(client):
class TestS3PresignClient:
@override_settings(
**{**PRESIGN_SETTINGS, "DJANGO_OUTPUT_S3_AWS_DEFAULT_REGION": ""},
DJANGO_OUTPUT_S3_AWS_PUBLIC_ENDPOINT_URL="",
)
def test_no_public_endpoint_and_no_region_returns_none(self):
# Without a region, SigV4 would have to guess one and break other regions.
assert get_s3_presign_client() is None
@override_settings(**PRESIGN_SETTINGS, DJANGO_OUTPUT_S3_AWS_PUBLIC_ENDPOINT_URL="")
def test_region_without_public_endpoint_signs_sigv4_on_the_regional_host(self):
# SSE-KMS objects reject the SigV2 URLs boto3 presigns by default, and the
# global host redirects for new buckets, which breaks a SigV4 signature.
url = urlparse(_presign(get_s3_presign_client()))
query = parse_qs(url.query)
assert url.netloc == "s3.eu-west-1.amazonaws.com"
assert url.path == "/output-bucket/tenant/scan/report.zip"
assert query["X-Amz-Algorithm"] == ["AWS4-HMAC-SHA256"]
assert "/eu-west-1/s3/aws4_request" in query["X-Amz-Credential"][0]
@override_settings(
**{
**PRESIGN_SETTINGS,
"DJANGO_OUTPUT_S3_AWS_ACCESS_KEY_ID": "",
"DJANGO_OUTPUT_S3_AWS_SECRET_ACCESS_KEY": "",
},
DJANGO_OUTPUT_S3_AWS_PUBLIC_ENDPOINT_URL="",
)
def test_region_without_static_keys_signs_with_the_default_chain(self, monkeypatch):
# An ECS task role reaches boto3 through the default chain, like the env here.
# A fresh default session keeps these keys from being cached for later tests.
monkeypatch.setattr(boto3, "DEFAULT_SESSION", None)
monkeypatch.setenv("AWS_ACCESS_KEY_ID", "role-access-key")
monkeypatch.setenv("AWS_SECRET_ACCESS_KEY", "role-secret-key")
monkeypatch.setenv("AWS_DEFAULT_REGION", "us-east-1")
query = parse_qs(urlparse(_presign(get_s3_presign_client())).query)
assert query["X-Amz-Credential"][0].startswith("role-access-key/")
assert "/eu-west-1/s3/aws4_request" in query["X-Amz-Credential"][0]
@override_settings(
**{**PRESIGN_SETTINGS, "DJANGO_OUTPUT_S3_AWS_DEFAULT_REGION": ""},
DJANGO_OUTPUT_S3_AWS_PUBLIC_ENDPOINT_URL="",
DJANGO_OUTPUT_S3_AWS_ENDPOINT_URL="http://minio:9000",
)
def test_internal_endpoint_without_public_endpoint_signs_against_it(self):
# No browser-reachable host was configured, so the internal one is the best
# available target instead of falling through to the real AWS host.
url = urlparse(_presign(get_s3_presign_client()))
assert url.netloc == "minio:9000"
assert url.path == "/output-bucket/tenant/scan/report.zip"
@override_settings(
**PRESIGN_SETTINGS,
DJANGO_OUTPUT_S3_AWS_PUBLIC_ENDPOINT_URL="https://storage.example.com",
DJANGO_OUTPUT_S3_AWS_ENDPOINT_URL="http://minio:9000",
)
def test_public_endpoint_wins_over_the_internal_endpoint(self):
url = urlparse(_presign(get_s3_presign_client()))
assert url.netloc == "storage.example.com"
def test_no_public_endpoint_returns_none(self):
assert get_s3_presign_client() is None
@override_settings(
**PRESIGN_SETTINGS,
+5 -178
View File
@@ -12,7 +12,6 @@ from api.db_router import MainRouter
from api.db_utils import rls_transaction
from api.exceptions import ProviderConnectionError, ProviderDeletedException
from api.models import (
AttackSurfaceOverview,
Finding,
MuteRule,
Provider,
@@ -5328,13 +5327,8 @@ class TestAggregateAttackSurface:
mock_queryset = MagicMock()
mock_queryset.values.return_value = mock_queryset
mock_queryset.annotate.return_value = [
{
"check_id": "check_internet_1",
"total": 10,
"failed": 3,
"muted_count": 1,
},
{"check_id": "check_secrets_1", "total": 5, "failed": 2, "muted_count": 0},
{"check_id": "check_internet_1", "total": 10, "failed": 3, "muted": 1},
{"check_id": "check_secrets_1", "total": 5, "failed": 2, "muted": 0},
]
ctx = MagicMock()
@@ -5383,7 +5377,7 @@ class TestAggregateAttackSurface:
mock_queryset = MagicMock()
mock_queryset.values.return_value = mock_queryset
mock_queryset.annotate.return_value = [
{"check_id": "check_internet_1", "total": 5, "failed": 1, "muted_count": 0},
{"check_id": "check_internet_1", "total": 5, "failed": 1, "muted": 0},
]
ctx = MagicMock()
@@ -5466,13 +5460,8 @@ class TestAggregateAttackSurface:
mock_queryset = MagicMock()
mock_queryset.values.return_value = mock_queryset
mock_queryset.annotate.return_value = [
{
"check_id": "check_internet_1",
"total": 10,
"failed": 3,
"muted_count": 1,
},
{"check_id": "check_internet_2", "total": 5, "failed": 2, "muted_count": 0},
{"check_id": "check_internet_1", "total": 10, "failed": 3, "muted": 1},
{"check_id": "check_internet_2", "total": 5, "failed": 2, "muted": 0},
]
ctx = MagicMock()
@@ -5493,62 +5482,6 @@ class TestAggregateAttackSurface:
assert overview.failed_findings == 5 # 3 + 2
assert overview.muted_failed_findings == 1 # 1 + 0
@patch("tasks.jobs.scan._get_attack_surface_mapping_from_provider")
def test_aggregate_attack_surface_counts_real_findings(
self, mock_get_mapping, tenants_fixture, scans_fixture
):
"""Run the aggregation query against real Finding rows.
The other tests mock the queryset, so they never execute the real
`annotate`. This one guards the row keys the query returns."""
tenant = tenants_fixture[0]
scan = scans_fixture[0]
mock_get_mapping.return_value = {
"privilege-escalation": {"check_privesc_1"},
"secrets": {"check_secrets_1"},
}
def create_finding(uid, check_id, status, muted):
Finding.objects.create(
tenant_id=tenant.id,
uid=uid,
scan=scan,
status=status,
status_extended="status extended",
impact=Severity.high,
severity=Severity.high,
raw_result={"status": status},
check_id=check_id,
check_metadata={"CheckId": check_id},
muted=muted,
first_seen_at="2024-01-02T00:00:00Z",
)
create_finding("privesc_fail", "check_privesc_1", Status.FAIL, False)
create_finding("privesc_fail_2", "check_privesc_1", Status.FAIL, False)
create_finding("privesc_fail_muted", "check_privesc_1", Status.FAIL, True)
create_finding("privesc_pass", "check_privesc_1", Status.PASS, False)
create_finding("secrets_pass_muted", "check_secrets_1", Status.PASS, True)
create_finding("unmapped_fail", "check_unmapped", Status.FAIL, False)
aggregate_attack_surface(str(tenant.id), str(scan.id))
overviews = {
overview.attack_surface_type: overview
for overview in AttackSurfaceOverview.objects.filter(
tenant_id=tenant.id, scan_id=scan.id
)
}
assert set(overviews) == {"privilege-escalation", "secrets"}
assert overviews["privilege-escalation"].total_findings == 4
assert overviews["privilege-escalation"].failed_findings == 2
assert overviews["privilege-escalation"].muted_failed_findings == 1
assert overviews["secrets"].total_findings == 1
assert overviews["secrets"].failed_findings == 0
assert overviews["secrets"].muted_failed_findings == 0
@patch("tasks.jobs.scan.Scan.all_objects.select_related")
@patch("tasks.jobs.scan.rls_transaction")
def test_aggregate_attack_surface_uses_select_related(
@@ -6034,112 +5967,6 @@ class TestResetEphemeralResourceFindingsCount:
resource2.refresh_from_db()
assert resource2.failed_findings_count == 5
def test_runs_when_newer_scan_is_not_full_scope(
self, tenants_fixture, scans_fixture, aws_provider, resources_fixture
):
"""A newer scoped scan must not block the full-scope scan's cleanup.
The race guard used to pick the newest COMPLETED scan of any scope
despite being named after full-scope ones, so a single scoped scan
landing after a complete one made the complete scan skip its cleanup
and leave ephemeral resources with a stale count permanently.
"""
from datetime import timedelta
tenant, *_ = tenants_fixture
scan1, *_ = scans_fixture
resource1, resource2, _ = resources_fixture
Resource.objects.filter(id=resource2.id).update(failed_findings_count=5)
self._make_scan_summary(tenant.id, scan1.id, resource1)
newer_completed_at = scan1.completed_at + timedelta(minutes=5)
Scan.objects.create(
name="Newer scoped scan",
provider=aws_provider,
trigger=Scan.TriggerChoices.MANUAL,
state=StateChoices.COMPLETED,
tenant_id=tenant.id,
started_at=newer_completed_at,
completed_at=newer_completed_at,
scanner_args={"checks": ["check1"]},
)
result = reset_ephemeral_resource_findings_count(
tenant_id=str(tenant.id), scan_id=str(scan1.id)
)
assert result["status"] == "completed"
resource2.refresh_from_db()
assert resource2.failed_findings_count == 0
def test_runs_when_many_newer_scans_are_not_full_scope(
self, tenants_fixture, scans_fixture, aws_provider, resources_fixture
):
"""The walk must not give up before it reaches the full-scope scan.
A fixed look-back window returned None once more newer scoped scans
had landed than it inspected, and skipped the cleanup exactly like the
original bug did with one.
"""
from datetime import timedelta
tenant, *_ = tenants_fixture
scan1, *_ = scans_fixture
resource1, resource2, _ = resources_fixture
Resource.objects.filter(id=resource2.id).update(failed_findings_count=5)
self._make_scan_summary(tenant.id, scan1.id, resource1)
for minutes in range(1, 41):
newer_completed_at = scan1.completed_at + timedelta(minutes=minutes)
Scan.objects.create(
name=f"Newer scoped scan {minutes}",
provider=aws_provider,
trigger=Scan.TriggerChoices.MANUAL,
state=StateChoices.COMPLETED,
tenant_id=tenant.id,
started_at=newer_completed_at,
completed_at=newer_completed_at,
scanner_args={"checks": ["check1"]},
)
result = reset_ephemeral_resource_findings_count(
tenant_id=str(tenant.id), scan_id=str(scan1.id)
)
assert result["status"] == "completed"
resource2.refresh_from_db()
assert resource2.failed_findings_count == 0
def test_runs_when_completed_at_is_null(
self, tenants_fixture, scans_fixture, aws_provider, resources_fixture
):
"""NULL `completed_at` used to make the reset never run at all.
The old guard filtered `completed_at__isnull=False`, so a provider
whose completed scans all had a NULL `completed_at` resolved the
"latest" scan to None, which never equals `scan.id`.
"""
tenant, *_ = tenants_fixture
scan1, *_ = scans_fixture
resource1, resource2, _ = resources_fixture
Scan.all_objects.filter(id=scan1.id).update(completed_at=None)
Resource.objects.filter(id=resource2.id).update(failed_findings_count=5)
self._make_scan_summary(tenant.id, scan1.id, resource1)
result = reset_ephemeral_resource_findings_count(
tenant_id=str(tenant.id), scan_id=str(scan1.id)
)
assert result["status"] == "completed"
resource2.refresh_from_db()
assert resource2.failed_findings_count == 0
def test_does_not_touch_other_providers_resources(
self, tenants_fixture, scans_fixture, aws_provider, resources_fixture
):
-79
View File
@@ -1,4 +1,3 @@
import json
import uuid
from contextlib import contextmanager
from datetime import UTC, datetime
@@ -15,7 +14,6 @@ from api.models import (
StateChoices,
Task,
)
from api.v1.serializers import TaskSerializer
from botocore.exceptions import ClientError
from celery import states
from django_celery_beat.models import IntervalSchedule, PeriodicTask
@@ -34,7 +32,6 @@ from tasks.tasks import (
_scan_tmp_output_directory,
check_integrations_task,
check_lighthouse_provider_connection_task,
create_scan_task_record,
generate_outputs_task,
mute_findings_in_latest_scans_task,
perform_attack_paths_scan_task,
@@ -3387,79 +3384,3 @@ class TestTaskTimeLimits:
"lighthouse-provider-connection-check",
):
assert celery_app.tasks[name].time_limit < default
@pytest.mark.django_db
class TestCreateScanTaskRecord:
"""`task_kwargs` is what a response built before the publish can report."""
def _scan(self, tenant, provider):
"""A manual scan, like the one `POST /api/v1/scans` creates."""
return Scan.objects.create(
tenant_id=tenant.id,
provider=provider,
name="Manual scan",
trigger=Scan.TriggerChoices.MANUAL,
state=StateChoices.AVAILABLE,
)
def _publish_kwargs(self, tenant, scan):
"""What `enqueue_scan_execution_on_commit` publishes for this scan."""
return {
"tenant_id": str(tenant.id),
"scan_id": str(scan.id),
"provider_id": str(scan.provider_id),
}
def _task_args(self, task):
"""Read the record back the way `TaskSerializer` does."""
return TaskSerializer(task).data["task_args"]
def test_the_stored_kwargs_are_the_ones_the_publish_would_send(
self, tenants_fixture, aws_provider
):
"""The 202 reports what is stored here, so it has to be the dispatch kwargs."""
tenant = tenants_fixture[0]
scan = self._scan(tenant, aws_provider)
task = create_scan_task_record(
tenant_id=str(tenant.id),
task_id=str(uuid.uuid4()),
task_kwargs=self._publish_kwargs(tenant, scan),
)
assert self._task_args(task) == {
"scan_id": str(scan.id),
"provider_id": str(aws_provider.id),
}
def test_a_record_created_without_kwargs_reports_none(self, tenants_fixture):
"""The argument is optional, so the other callers keep their behaviour."""
task = create_scan_task_record(
tenant_id=str(tenants_fixture[0].id),
task_id=str(uuid.uuid4()),
)
assert self._task_args(task) == {}
def test_the_publish_can_overwrite_the_stored_kwargs(
self, tenants_fixture, aws_provider
):
"""django-celery-results stores a Python repr; both must decode alike."""
tenant = tenants_fixture[0]
scan = self._scan(tenant, aws_provider)
task_id = str(uuid.uuid4())
kwargs = self._publish_kwargs(tenant, scan)
task = create_scan_task_record(
tenant_id=str(tenant.id), task_id=task_id, task_kwargs=kwargs
)
before = self._task_args(task)
# What `before_task_publish` writes once the task reaches the broker.
task_result = TaskResult.objects.get(task_id=task_id)
task_result.task_kwargs = json.dumps(repr(kwargs))
task_result.save(update_fields=["task_kwargs"])
task.refresh_from_db()
assert self._task_args(task) == before
Generated
+9 -21
View File
@@ -291,7 +291,7 @@ constraints = [
{ name = "pydantic-core", specifier = "==2.41.5" },
{ name = "pygithub", specifier = "==2.8.0" },
{ name = "pygments", specifier = "==2.20.0" },
{ name = "pyjwt", specifier = "==2.14.0" },
{ name = "pyjwt", specifier = "==2.13.0" },
{ name = "pylint", specifier = "==3.2.5" },
{ name = "pymsalruntime", specifier = "==0.18.1" },
{ name = "pynacl", specifier = "==1.6.2" },
@@ -387,7 +387,7 @@ overrides = [
{ name = "microsoft-kiota-serialization-multipart", specifier = "==1.9.10" },
{ name = "microsoft-kiota-serialization-text", specifier = "==1.9.10" },
{ name = "okta", specifier = "==3.4.2" },
{ name = "pyjwt", extras = ["crypto"], specifier = "==2.14.0" },
{ name = "pyjwt", extras = ["crypto"], specifier = "==2.13.0" },
]
[[package]]
@@ -3501,17 +3501,6 @@ wheels = [
{ url = "https://files.pythonhosted.org/packages/18/7d/721f162c46e3de604a73674223bf6c6bc6cf7ade25b3751a71288f4dd122/huaweicloudsdkrds-3.1.204-py3-none-any.whl", hash = "sha256:a790b5b3c457a608e5679c101f463b4d037dd9a8a66f6e46144a9e5a4b37780f", size = 1626906, upload-time = "2026-07-09T09:04:06.936Z" },
]
[[package]]
name = "huaweicloudsdksmn"
version = "3.1.204"
source = { registry = "https://pypi.org/simple" }
dependencies = [
{ name = "huaweicloudsdkcore" },
]
wheels = [
{ url = "https://files.pythonhosted.org/packages/e9/5d/c0de47d011f1932c9c0ddf669c238d9fad01653d438d38203703526799d7/huaweicloudsdksmn-3.1.204-py3-none-any.whl", hash = "sha256:b0818ea9293e27458c8fa1d2da021c98006cbf9ce01cf17a0f1df598c4da3a6c", size = 323674, upload-time = "2026-07-09T09:04:24.804Z" },
]
[[package]]
name = "huaweicloudsdkvpc"
version = "3.1.204"
@@ -4846,8 +4835,8 @@ wheels = [
[[package]]
name = "prowler"
version = "5.44.0"
source = { git = "https://github.com/prowler-cloud/prowler.git?rev=v5.44#60b936005cc109c611ad8b7f7c41cadb586fb4b6" }
version = "5.41.0"
source = { git = "https://github.com/prowler-cloud/prowler.git?rev=master#f05a490cd74a2c0f11a5d66d8ce29d03fa5c64a2" }
dependencies = [
{ name = "alibabacloud-actiontrail20200706" },
{ name = "alibabacloud-credentials" },
@@ -4911,7 +4900,6 @@ dependencies = [
{ name = "huaweicloudsdkkms" },
{ name = "huaweicloudsdkobs" },
{ name = "huaweicloudsdkrds" },
{ name = "huaweicloudsdksmn" },
{ name = "huaweicloudsdkvpc" },
{ name = "huaweicloudsdkwaf" },
{ name = "jsonschema" },
@@ -4950,7 +4938,7 @@ dependencies = [
[[package]]
name = "prowler-api"
version = "1.45.1"
version = "1.44.0"
source = { virtual = "." }
dependencies = [
{ name = "cartography" },
@@ -5050,7 +5038,7 @@ requires-dist = [
{ name = "matplotlib", specifier = "==3.10.8" },
{ name = "neo4j", specifier = "==6.1.0" },
{ name = "openai", specifier = "==1.109.1" },
{ name = "prowler", git = "https://github.com/prowler-cloud/prowler.git?rev=v5.44" },
{ name = "prowler", git = "https://github.com/prowler-cloud/prowler.git?rev=master" },
{ name = "psycopg2-binary", specifier = "==2.9.9" },
{ name = "pytest-celery", extras = ["redis"], specifier = "==1.3.0" },
{ name = "reportlab", specifier = "==4.4.10" },
@@ -5344,11 +5332,11 @@ wheels = [
[[package]]
name = "pyjwt"
version = "2.14.0"
version = "2.13.0"
source = { registry = "https://pypi.org/simple" }
sdist = { url = "https://files.pythonhosted.org/packages/af/c3/8a3b59c25070cc61dc517fbdfa5dc0904670c96f605cc69759dc09166b99/pyjwt-2.14.0.tar.gz", hash = "sha256:77283c83fb56ecf566a886c757a714bc83668e38156de2cce8263302f42e0b86", size = 113177, upload-time = "2026-09-11T13:11:54.638Z" }
sdist = { url = "https://files.pythonhosted.org/packages/3b/81/58d0ac84e1ef3a3843791d6954d94c0b33d526c75eeb1efbce9d0a4c4077/pyjwt-2.13.0.tar.gz", hash = "sha256:41571c89ca91598c79e8ef18a2d07367d4810fbbd6f637794879baf1b7703423", size = 107515, upload-time = "2026-05-21T19:54:36.618Z" }
wheels = [
{ url = "https://files.pythonhosted.org/packages/9c/97/672cb32ce0dfea44b740cb7b4f97038463b9cf7c0ead1aacf595572851d6/pyjwt-2.14.0-py3-none-any.whl", hash = "sha256:ad0cef71c756a56e74863c2919cf0985f72decbcfcb550ee2f422e7c62b5eedc", size = 32896, upload-time = "2026-09-11T13:11:53.409Z" },
{ url = "https://files.pythonhosted.org/packages/a3/5e/ecf12fdb62546d64385c158514e9b2b671f7832108ef2ecd2020ce0af2d1/pyjwt-2.13.0-py3-none-any.whl", hash = "sha256:66adcc2aff09b3f1bbd95fc1e1577df8ac8723c978552fd43304c8a290ac5728", size = 31274, upload-time = "2026-05-21T19:54:35.362Z" },
]
[package.optional-dependencies]
-8
View File
@@ -15,7 +15,6 @@ services:
api:
hostname: "prowler-api"
image: prowlercloud/prowler-api:${PROWLER_API_VERSION:-stable}
restart: unless-stopped
env_file:
- path: .env
required: false
@@ -45,7 +44,6 @@ services:
ui:
image: prowlercloud/prowler-ui:${PROWLER_UI_VERSION:-stable}
restart: unless-stopped
env_file:
- path: .env
required: false
@@ -63,7 +61,6 @@ services:
postgres:
image: postgres:16-alpine@sha256:57c72fd2a128e416c7fcc499958864df5301e940bca0a56f58fddf30ffc07777
restart: unless-stopped
hostname: "postgres-db"
volumes:
- ./_data/postgres:/var/lib/postgresql/data
@@ -84,7 +81,6 @@ services:
valkey:
image: valkey/valkey:8-alpine@sha256:a038175878d66b9d274fbf8be73c0305e93798b83917647f167e18cef3c71eec
restart: unless-stopped
hostname: "valkey"
volumes:
- ./_data/valkey:/data
@@ -101,7 +97,6 @@ services:
neo4j:
image: graphstack/dozerdb:5.26.27.0@sha256:9b54d6b3a98a76c00bd23e8e78d8c82081ff168162aebd47b25c234e092cb0a0
restart: unless-stopped
hostname: "neo4j"
volumes:
- ./_data/neo4j:/data
@@ -134,7 +129,6 @@ services:
worker:
image: prowlercloud/prowler-api:${PROWLER_API_VERSION:-stable}
restart: unless-stopped
# Give Celery soft shutdown time to drain/re-queue in-flight tasks on stop.
stop_grace_period: 120s
env_file:
@@ -155,7 +149,6 @@ services:
worker-beat:
image: prowlercloud/prowler-api:${PROWLER_API_VERSION:-stable}
restart: unless-stopped
env_file:
- path: ./.env
required: false
@@ -172,7 +165,6 @@ services:
mcp-server:
image: prowlercloud/prowler-mcp:${PROWLER_MCP_VERSION:-stable}
restart: unless-stopped
environment:
- PROWLER_MCP_TRANSPORT_MODE=http
env_file:
-67
View File
@@ -4,73 +4,6 @@ description: "New features and improvements in each Prowler release"
rss: true
---
<Update label="v5.43.0" description="September 21, 2026">
### 🏛️ Compliance — FedRAMP 20x Consolidated Rules 2026
The FedRAMP 20x Phase One pilot frameworks (`fedramp_20x_ksi_low_aws`, `fedramp_20x_ksi_low_azure` and `fedramp_20x_ksi_low_gcp`) are replaced by two universal frameworks built from the FedRAMP Consolidated Rules for 2026, each covering AWS, Azure, GCP, Kubernetes and Microsoft 365 from a single definition:
- **FedRAMP 20x KSI** (`fedramp_20x_ksi_2026`): the 46 Key Security Indicators across 10 themes. There is one indicator catalog for every class instead of a separate Low fork; each indicator carries its class applicability and NIST SP 800-53 controls.
- **FedRAMP 20x Class C FRR** (`fedramp_20x_frr_class_c_2026`): the 158 FedRAMP Rules of the Class C ruleset that bind cloud service providers. Most are program obligations (reports, notifications, certification package) and stay manual; checks are mapped only where they evidence part of the rule text. Configurable checks carry configuration requirements, so a relaxed `audit_config` cannot turn a requirement green.
Automation or stored results that reference the pilot framework IDs need to move to `fedramp_20x_ksi_2026`. In Prowler App, both frameworks show the per-provider breakdown in the cross-provider compliance view and can be downloaded as OCSF.
Read more in the [Compliance documentation](https://docs.prowler.com/user-guide/compliance/tutorials/compliance).
### 🔎 AWS — Inspector Coverage, CISA KEV and FIPS Checks
Seven new AWS checks back the vulnerability detection and cryptography rules of FedRAMP 20x Class C:
- `inspector2_coverage_scan_status_active` and `inspector2_coverage_recently_scanned` report resources Amazon Inspector is not scanning, or last scanned more than `inspector2_max_days_since_last_scan` days ago (default 3).
- `inspector2_active_findings_no_known_exploited_vulnerabilities` and `inspector2_active_findings_kev_within_due_date` report active findings whose CVE is in the CISA Known Exploited Vulnerabilities catalog, and those still open past the CISA due date. The KEV data comes from Inspector itself through `inspector2:BatchGetFindingDetails`, so no external feed is needed.
- `inspector2_active_findings_within_max_age` reports active findings first observed more than `inspector2_active_finding_max_age_days` days ago (default 192).
- `elbv2_listener_fips_tls_enabled` and `transfer_server_fips_security_policy_enabled` report HTTPS/TLS load balancer listeners and Transfer Family servers without a FIPS security policy.
`inspector2:BatchGetFindingDetails` is not part of `SecurityAudit`, so it is now included in the Prowler additions policy and the CloudFormation scan role. Without it, the KEV checks report `MANUAL` naming the missing permission instead of a false `FAIL`.
Explore all AWS checks at [Prowler Hub](https://hub.prowler.com/check?provider=aws).
### ☁️ AWS — Partition Bootstrap Failover
When `PROWLER_AWS_PARTITION` is set, the bootstrap STS calls (validating credentials, assuming a role and getting an MFA session token) now try up to two more regions of the partition if the first one cannot be reached. A GovCloud host whose configured region belongs to another partition was still sent to `us-gov-east-1`, and on a network that routes only to `us-gov-west-1` the connection check and the scan failed on perfectly valid credentials. Only connection errors and timeouts move on to the next region; credential errors are reported from the first one as before. Later STS calls reuse the region that answered, and nothing changes when `PROWLER_AWS_PARTITION` is unset.
Read more in the [AWS Regions and Partitions documentation](https://docs.prowler.com/user-guide/providers/aws/regions-and-partitions).
### 🌐 Azure — Sovereign Cloud Endpoints for Defender and Key Vault
Defender security contacts and Key Vault key rotation policies now use the endpoints of the cloud selected with `--azure-region` instead of the hardcoded `management.azure.com` and `vault.azure.net` hosts, so both work on `AzureUSGovernment` and `AzureChinaCloud`. Key Vault clients are built from the vault URI that Azure returns for each vault.
Read more in the [Azure non-default cloud documentation](https://docs.prowler.com/user-guide/providers/azure/use-non-default-cloud).
### ✉️ Invitations — Expired Invitations No Longer Block Re-Invites
A pending invitation past its expiry date is now reported as expired, and inviting the same email again marks it as expired and creates the new invitation instead of returning a generic error. The Invitations table disables Edit and Revoke on expired and revoked invitations, and `filter[state__in]` on the invitations endpoint no longer returns a server error.
In Prowler Cloud, new organizations are offered an **Invite your team** step once the first provider is connected, and Prowler Private Cloud deployments can set `UI_SELF_REGISTRATION_ENABLED=false` to make sign-up invitation-only.
Read more in the [Invitations documentation](https://docs.prowler.com/user-guide/tutorials/prowler-app-rbac#invitations).
### 📄 Reports — Downloads on Self-Hosted Storage
Report downloads no longer depend on the storage host being the same inside and outside the container network. `DJANGO_OUTPUT_S3_AWS_PUBLIC_ENDPOINT_URL` signs the download URL against a browser-reachable host, so a deployment whose object storage answers only on an internal address serves the file instead of a link the browser cannot open. Leaving `DJANGO_OUTPUT_S3_AWS_DEFAULT_REGION` unset, common on S3-compatible storage with no meaningful region, no longer makes the download fail with a server error.
### 🔍 Checks
- **Huawei Cloud:** new `smn_topic_subscriptions` check reports SMN topics without any subscription.
- **Cloudflare:** the API token links in the provider wizard request the SSL and Certificates, Bot Management and Zone WAF read permissions the checks need, so a token created from the wizard no longer produces failures on permissions it was never granted.
- **Microsoft 365:** five Defender malware, anti-phishing and inbound anti-spam checks no longer fail with `KeyError` on tenants that use the Standard or Strict preset security policies, which dropped every finding of those checks. Preset policies are covered by `defender_strict_preset_security_policy_enabled`.
- **Google Workspace:** `security_2sv_enforced` reports domain-wide 2-Step Verification failures as `FAIL` even when every failing setting is overridden for a group or organizational unit.
Explore all checks at [Prowler Hub](https://hub.prowler.com/check).
### 🔐 Security Updates
- `libsqlite3-0`, `gzip`, `perl-base`, `libssh2-1t64` and `libpcre2-8-0` upgraded in the SDK and API container images, patching high-severity Debian CVEs.
- PowerShell upgraded to 7.5.11 in the SDK and API container images, bundling .NET runtime 9.0.20 and patching CVE-2026-62901.
- `anyio` upgraded to 4.14.2 in the SDK, the API and the MCP Server, patching CVE-2026-63374.
See the [full release notes on GitHub](https://github.com/prowler-cloud/prowler/releases/tag/5.43.0) for the complete list of changes.
</Update>
<Update label="v5.42.0" description="September 11, 2026">
### ☁️ AWS — ISO Partitions
@@ -47,8 +47,6 @@ The former build-time variables map to the new runtime variables as follows:
Roll out Registry only after the Registry backend dependency is deployed, intended roles have `manage_registry`, and acceptance with real credentials has exercised installation, provider account creation, credentials, connection, and scan launch. Deploy the UI with `UI_REGISTRY_ENABLED` unset or `"false"`; set it to `"true"` only in the prepared process environment, then restart or otherwise apply the environment update required by the platform. A Registry key must belong to the configured Registry environment; a production key does not authenticate against a development Registry.
The Registry links read `PROWLER_REGISTRY_INDEX_URL`, the same base URL the backend installs artifacts from. The Registry page and the credential banner link to it, and its origin is allowed for images. `UI_REGISTRY_MEDIA_URL` adds the Registry media service origin so artifact logos load. When `PROWLER_REGISTRY_INDEX_URL` is unset or invalid, the links are hidden instead of pointing to the public Prowler Registry, which keeps private and air-gapped deployments from sending users to an unreachable host. `UI_REGISTRY_URL` is no longer read.
The catalog displays all artifacts, including built-ins and packages containing only checks or compliance frameworks. Only external provider artifacts support Add. After confirmed installation, open Providers and select the option labeled Registry to configure an account. Creating accounts and running scans also require the corresponding provider and scan permissions. Removing an artifact keeps existing provider accounts, but future connections or scans can fail until the artifact is installed again.
To roll back, set `UI_REGISTRY_ENABLED=false` or remove it and apply the environment update. Proxy, page, and action checks deny on their next request. Navigation refreshes from server-authorized access when the page is requested again. Rollback does not delete Registry credentials, tenant artifact records, or provider accounts.
@@ -128,8 +128,8 @@ To update the environment file:
Edit the `.env` file and change version values:
```env
PROWLER_UI_VERSION="5.43.0"
PROWLER_API_VERSION="5.43.0"
PROWLER_UI_VERSION="5.42.0"
PROWLER_API_VERSION="5.42.0"
```
<Note>
+1 -1
View File
@@ -71,7 +71,7 @@ Prowler supports a wide range of providers organized by category:
| [LLM](/user-guide/providers/llm/getting-started-llm) | Official | Models | CLI |
| [M365](/user-guide/providers/microsoft365/getting-started-m365) | Official | Tenants | UI, API, CLI |
| [MongoDB Atlas](/user-guide/providers/mongodbatlas/getting-started-mongodbatlas) | Official | Organizations | UI, API, CLI |
| [Okta](/user-guide/providers/okta/getting-started-okta) | Official | Organizations | UI, API, CLI |
| [Okta](/user-guide/providers/okta/getting-started-okta) | Official | Organizations | CLI |
| [Vercel](/user-guide/providers/vercel/getting-started-vercel) | Official | Teams / Projects | UI, API, CLI |
### Kubernetes
@@ -29,22 +29,6 @@ Boto3 defaults both timeouts to 60 seconds. In networks with restricted egress (
</Note>
## Retries Configuration
<VersionBadge version="5.44.0" />
The number of retries is set with `--aws-retries-max-attempts`, where `0` disables retries. It can also be set through an environment variable, which is the way to tune it in Prowler Cloud and other deployments without a CLI:
```console
export PROWLER_AWS_BOTO3_RETRIES_MAX_ATTEMPTS=0
```
The CLI flag takes precedence over the environment variable. The value must be a non-negative integer; when neither is set, Prowler uses 3 retries.
<Warning>
The environment variable is process-wide: it applies to every AWS provider built in the process where it is set, not only to a connection check. A scan started in that same process picks it up too. Boto3's Standard retry mode, which Prowler uses, also retries service-side throttling responses (see the errors listed below), so `0` disables retries for those as well. On a large account a scan can hit throttling under normal load, and with retries disabled that throttling becomes a hard failure instead of a retried call. Set the variable only on the processes that run connection checks; leave scan workers on the default, or raise their retry count instead of lowering it.
</Warning>
## Retry Behavior Overview
Boto3's Standard retry mode includes the following mechanisms:
@@ -12,7 +12,7 @@ The following steps apply to Prowler Cloud and Prowler Local Server.
1. Sign in to the [OCI Console](https://cloud.oracle.com/) and open **Tenancy Details** to copy the Tenancy OCID.
2. Go to **Identity & Security** → **Users**, select the principal that owns the API key, and copy the **User OCID**.
3. Generate or locate the API key fingerprint and private key for that user. Follow the [Config File Authentication steps](/user-guide/providers/oci/authentication#config-file-authentication-manual-api-key-setup) to create or rotate the key pair and copy the fingerprint.
4. In **Tenancy Details**, note the **Home Region** identifier (for example, `me-abudhabi-1`). Any other region the tenancy is subscribed to also works.
4. Note the **Region** identifier to scan (for example, `us-ashburn-1`).
### Step 2: Access Prowler Cloud
1. Navigate to [Prowler Cloud](https://cloud.prowler.com/) or launch [Prowler Local Server](/user-guide/tutorials/prowler-app).
@@ -26,18 +26,12 @@ Prowler Cloud connects to OCI with API key credentials. Provide:
- **User OCID** for the API key owner
- **Fingerprint** of the API key
- **Home Region**: select it from the list (for example, `me-abudhabi-1`)
- **Region** (for example, `us-ashburn-1`)
- **Private Key Content** (paste the full PEM value)
- **Passphrase (Optional)** if the private key is encrypted
Select **Next**, then **Launch Scan** to validate the connection and start the first OCI scan. The private key content is encoded for secure transmission.
<Note>
The home region is used only to validate the credentials and discover the regions the tenancy is subscribed to. It does not limit the scan: Prowler audits every subscribed region. OCI Identity and Access Management (IAM) only answers in subscribed regions, so a tenancy that is not subscribed to the selected region cannot be validated.
</Note>
Providers created without a region keep using `us-ashburn-1` for validation. If such a provider fails with `401 NotAuthenticated`, update its credentials and select the home region.
![Add OCI API Key Credentials](./images/oci-add-api-key-credentials.png)
---
@@ -340,11 +334,6 @@ prowler oci \
#### Region Issues
**Error: "OCI credential validation failed" with `401 NotAuthenticated` on `get_tenancy`**
- The tenancy is not subscribed to the region used for validation (by default `us-ashburn-1`)
- In Prowler Cloud or Prowler Local Server, update the provider credentials and select the tenancy home region
- In Prowler CLI, set `region` in `~/.oci/config` to the home region. All subscribed regions are still scanned; `--region` also works but limits the scan to that region
**Error: "Invalid region"**
- Check available regions: `prowler oci --list-regions`
- Verify your tenancy is subscribed to the region
@@ -148,12 +148,6 @@ New roles have no provider visibility by default. Assign at least one Provider G
Integrations follow the visibility of the providers attached to them: a role can see an integration when it can access at least one of its providers, and only the providers visible to that role are listed on the integration. Editing or deleting an integration attached to providers outside the visibility of the role is not allowed. Integrations that are not attached to any provider, such as Jira, are tenant-wide and remain available to every role with the **Manage Integrations** permission.
#### Task Visibility and Revocation
<VersionBadge version="5.44.0" />
Background tasks, such as provider deletions, connection checks and scans, follow the visibility of the provider they belong to: a role can see a task when it can access its provider. Tasks that carry no provider reference are treated as tenant-wide and are visible to every role. Revoking a pending task requires the same permission as the operation that queued it, for example **Manage Scans** for a scan. Provider deletions cannot be revoked.
#### Creating a Provider Group
Follow these steps to create a provider group in your account:
@@ -89,10 +89,6 @@ After adding your cloud account credentials, click the `Check connection` button
<img src="/images/test-connection-button.png" alt="Test Connection" width="700" />
<Note>
For a single AWS account, Prowler tests the connection as part of the `Connect account` step, so the wizard moves straight to launching the scan.
</Note>
## Step 6: Scan Started
After the connection check succeeds, save the provider and start your first scan with the `Launch Scan` button. The `Scans` section shows the scan in progress:
-8
View File
@@ -4,14 +4,6 @@ All notable changes to the **Prowler MCP Server** are documented in this file.
<!-- changelog: release notes start -->
## [0.12.2] (Prowler v5.43.0)
### 🔐 Security
- Bumped `anyio` to 4.14.2 to resolve CVE-2026-63374 [(#12848)](https://github.com/prowler-cloud/prowler/pull/12848)
---
## [0.12.1] (Prowler v5.42.0)
### 🔐 Security
@@ -0,0 +1 @@
Bumped `anyio` to 4.14.2 to resolve CVE-2026-63374
@@ -1 +0,0 @@
`pyjwt` from 2.13.0 to 2.14.0, patching GHSA-ffc3-869f-jxw9
-1
View File
@@ -81,6 +81,5 @@ constraint-dependencies = [
"cryptography==50.0.0",
"joserfc==1.6.8",
"mcp==1.28.1",
"pyjwt==2.14.0",
"python-multipart==0.0.30"
]
+5 -6
View File
@@ -13,7 +13,6 @@ constraints = [
{ name = "cryptography", specifier = "==50.0.0" },
{ name = "joserfc", specifier = "==1.6.8" },
{ name = "mcp", specifier = "==1.28.1" },
{ name = "pyjwt", specifier = "==2.14.0" },
{ name = "python-multipart", specifier = "==0.0.30" },
]
@@ -978,11 +977,11 @@ wheels = [
[[package]]
name = "pyjwt"
version = "2.14.0"
version = "2.13.0"
source = { registry = "https://pypi.org/simple" }
sdist = { url = "https://files.pythonhosted.org/packages/af/c3/8a3b59c25070cc61dc517fbdfa5dc0904670c96f605cc69759dc09166b99/pyjwt-2.14.0.tar.gz", hash = "sha256:77283c83fb56ecf566a886c757a714bc83668e38156de2cce8263302f42e0b86", size = 113177, upload-time = "2026-09-11T13:11:54.638Z" }
sdist = { url = "https://files.pythonhosted.org/packages/3b/81/58d0ac84e1ef3a3843791d6954d94c0b33d526c75eeb1efbce9d0a4c4077/pyjwt-2.13.0.tar.gz", hash = "sha256:41571c89ca91598c79e8ef18a2d07367d4810fbbd6f637794879baf1b7703423", size = 107515, upload-time = "2026-05-21T19:54:36.618Z" }
wheels = [
{ url = "https://files.pythonhosted.org/packages/9c/97/672cb32ce0dfea44b740cb7b4f97038463b9cf7c0ead1aacf595572851d6/pyjwt-2.14.0-py3-none-any.whl", hash = "sha256:ad0cef71c756a56e74863c2919cf0985f72decbcfcb550ee2f422e7c62b5eedc", size = 32896, upload-time = "2026-09-11T13:11:53.409Z" },
{ url = "https://files.pythonhosted.org/packages/a3/5e/ecf12fdb62546d64385c158514e9b2b671f7832108ef2ecd2020ce0af2d1/pyjwt-2.13.0-py3-none-any.whl", hash = "sha256:66adcc2aff09b3f1bbd95fc1e1577df8ac8723c978552fd43304c8a290ac5728", size = 31274, upload-time = "2026-05-21T19:54:35.362Z" },
]
[package.optional-dependencies]
@@ -1293,8 +1292,8 @@ name = "secretstorage"
version = "3.5.0"
source = { registry = "https://pypi.org/simple" }
dependencies = [
{ name = "cryptography", marker = "sys_platform != 'win32'" },
{ name = "jeepney", marker = "sys_platform != 'win32'" },
{ name = "cryptography" },
{ name = "jeepney" },
]
sdist = { url = "https://files.pythonhosted.org/packages/1c/03/e834bcd866f2f8a49a85eaff47340affa3bfa391ee9912a952a1faa68c7b/secretstorage-3.5.0.tar.gz", hash = "sha256:f04b8e4689cbce351744d5537bf6b1329c6fc68f91fa666f60a380edddcd11be", size = 19884, upload-time = "2025-11-23T19:02:53.191Z" }
wheels = [
@@ -0,0 +1,144 @@
AWSTemplateFormatVersion: "2010-09-09"
Description: |
Creates the read-only ProwlerScan IAM Role for a single AWS account so Prowler Cloud
can scan it. Trimmed-down version of prowler-scan-role.yml for the quick onboarding
flow: the only input is the External ID, everything else is fixed for Prowler Cloud.
For self-hosted Prowler, AWS Organizations or the S3 integration use prowler-scan-role.yml.
Parameters:
ExternalId:
Description: |
External ID that Prowler Cloud will use to assume the ProwlerScan IAM Role. Pre-filled by Prowler, do not edit.
Type: String
MinLength: 1
AllowedPattern: ".+"
ConstraintDescription: "ExternalId must not be empty."
Mappings:
ProwlerCloud:
Principal:
AccountId: "232136659152"
IAMPrincipal: "role/prowler*"
Resources:
ProwlerScan:
Type: AWS::IAM::Role
Properties:
RoleName: ProwlerScan
AssumeRolePolicyDocument:
Version: "2012-10-17"
Statement:
- Effect: Allow
Principal:
AWS: !Sub
- "arn:${AWS::Partition}:iam::${AccountId}:root"
- AccountId: !FindInMap [ProwlerCloud, Principal, AccountId]
Action: "sts:AssumeRole"
Condition:
StringEquals:
"sts:ExternalId": !Ref ExternalId
StringLike:
"aws:PrincipalArn": !Sub
- "arn:${AWS::Partition}:iam::${AccountId}:${IAMPrincipal}"
- AccountId: !FindInMap [ProwlerCloud, Principal, AccountId]
IAMPrincipal: !FindInMap [ProwlerCloud, Principal, IAMPrincipal]
MaxSessionDuration: 3600
ManagedPolicyArns:
- !Sub "arn:${AWS::Partition}:iam::aws:policy/SecurityAudit"
- !Sub "arn:${AWS::Partition}:iam::aws:policy/job-function/ViewOnlyAccess"
Policies:
- PolicyName: ProwlerScan
PolicyDocument:
Version: "2012-10-17"
Statement:
- Sid: AllowMoreReadOnly
Effect: Allow
Action:
- "account:Get*"
- "amplify:ListApps"
- "amplify:ListBranches"
- "appstream:Describe*"
- "appstream:List*"
- "backup:List*"
- "backup:Get*"
- "bedrock:List*"
- "bedrock:Get*"
- "cloudtrail:GetInsightSelectors"
- "codeartifact:List*"
- "codebuild:BatchGet*"
- "codebuild:ListReportGroups"
- "cognito-idp:GetUserPoolMfaConfig"
- "datapipeline:DescribePipelines"
- "datapipeline:GetPipelineDefinition"
- "datapipeline:ListPipelines"
- "dlm:Get*"
- "drs:Describe*"
- "ds:Get*"
- "ds:Describe*"
- "ds:List*"
- "dynamodb:GetResourcePolicy"
- "ec2:GetEbsEncryptionByDefault"
- "ec2:GetSnapshotBlockPublicAccessState"
- "ec2:GetInstanceMetadataDefaults"
- "ecr:Describe*"
- "ecr:GetRegistryScanningConfiguration"
- "ecr:BatchGetImage"
- "ecr:GetDownloadUrlForLayer"
- "elasticfilesystem:DescribeBackupPolicy"
- "glue:GetConnections"
- "glue:GetSecurityConfiguration*"
- "glue:SearchTables"
- "glue:GetMLTransforms"
- "inspector2:BatchGetFindingDetails"
- "lambda:GetFunction*"
- "logs:FilterLogEvents"
- "lightsail:GetRelationalDatabases"
- "macie2:GetMacieSession"
- "macie2:GetAutomatedDiscoveryConfiguration"
- "rolesanywhere:ListProfiles"
- "rolesanywhere:ListTagsForResource"
- "rolesanywhere:ListTrustAnchors"
- "s3:GetAccountPublicAccessBlock"
- "shield:DescribeProtection"
- "shield:GetSubscriptionState"
- "securityhub:GetFindings"
- "servicecatalog:Describe*"
- "servicecatalog:List*"
- "ssm:GetDocument"
- "ssm-incidents:List*"
- "states:ListTagsForResource"
- "support:Describe*"
- "tag:GetTagKeys"
- "wellarchitected:List*"
Resource: "*"
- Sid: AllowSecurityHubImportFindings
Effect: Allow
Action:
- "securityhub:BatchImportFindings"
Resource: "*"
- Sid: AllowAPIGatewayReadOnly
Effect: Allow
Action:
- "apigateway:GET"
Resource:
- !Sub "arn:${AWS::Partition}:apigateway:*::/restapis/*"
- !Sub "arn:${AWS::Partition}:apigateway:*::/apis/*"
- !Sub "arn:${AWS::Partition}:apigateway:*::/domainnames"
- !Sub "arn:${AWS::Partition}:apigateway:*::/domainnames/*"
Tags:
- Key: "Service"
Value: "https://prowler.com"
- Key: "Support"
Value: "support@prowler.com"
- Key: "CloudFormation"
Value: "true"
- Key: "Name"
Value: "ProwlerScan"
Outputs:
ProwlerScanRoleArn:
Description: "ARN of the ProwlerScan IAM Role. Paste it into Prowler."
Value: !GetAtt ProwlerScan.Arn
Export:
Name: !Sub "${AWS::StackName}-ProwlerScanRoleArn"
-28
View File
@@ -4,34 +4,6 @@ All notable changes to the **Prowler SDK** are documented in this file.
<!-- changelog: release notes start -->
## [5.43.0] (Prowler v5.43.0)
### 🚀 Added
- `FedRAMP-20x-KSI` universal compliance framework (`fedramp_20x_ksi_2026`) with the 46 Key Security Indicators from the FedRAMP Consolidated Rules 2026 mapped for AWS, Azure, GCP, Kubernetes and M365 [(#11701)](https://github.com/prowler-cloud/prowler/pull/11701)
- `smn_topic_subscriptions` check for Huawei Cloud provider: SMN topics have at least one subscription configured [(#12186)](https://github.com/prowler-cloud/prowler/pull/12186)
- `inspector2_coverage_scan_status_active`, `inspector2_coverage_recently_scanned`, `inspector2_active_findings_no_known_exploited_vulnerabilities`, `inspector2_active_findings_kev_within_due_date`, `inspector2_active_findings_within_max_age`, `elbv2_listener_fips_tls_enabled` and `transfer_server_fips_security_policy_enabled` checks for AWS provider, covering FedRAMP 20x Class C vulnerability detection, CISA KEV remediation and FIPS cryptography rules; the KEV checks require `inspector2:BatchGetFindingDetails`, now in the Prowler additions policy [(#12808)](https://github.com/prowler-cloud/prowler/pull/12808)
- `FedRAMP-20x-FRR-Class-C` universal compliance framework (`fedramp_20x_frr_class_c_2026`) with the 158 provider rules of the FedRAMP 20x Class C ruleset from the FedRAMP Consolidated Rules 2026 for AWS, Azure, GCP, Kubernetes and M365 [(#12808)](https://github.com/prowler-cloud/prowler/pull/12808)
### 🔄 Changed
- FedRAMP 20x Phase One pilot frameworks `fedramp_20x_ksi_low_aws`, `fedramp_20x_ksi_low_azure` and `fedramp_20x_ksi_low_gcp` replaced by `fedramp_20x_ksi_2026` [(#12855)](https://github.com/prowler-cloud/prowler/pull/12855)
### 🐞 Fixed
- `security_2sv_enforced` reports domain-wide 2-Step Verification failures as FAIL even when every failing setting is overridden for a group or organizational unit [(#12700)](https://github.com/prowler-cloud/prowler/pull/12700)
- Bootstrap STS calls now try up to two more regions of the partition declared in `PROWLER_AWS_PARTITION` when the first one cannot be reached, so a deployment that routes to only one region of its partition no longer fails on an endpoint it has no path to. This covers validating credentials, assuming a role and getting an MFA session token [(#12799)](https://github.com/prowler-cloud/prowler/pull/12799)
- `KeyError` in M365 Defender malware, anti-phishing and inbound anti-spam checks when the tenant has Standard or Strict preset security policies [(#12809)](https://github.com/prowler-cloud/prowler/pull/12809)
- Azure Defender security contacts and Key Vault key rotation policies now use the endpoints of the selected cloud (`--azure-region`) instead of the hardcoded `management.azure.com` and `vault.azure.net` hosts, so both work on `AzureUSGovernment` and `AzureChinaCloud` [(#12813)](https://github.com/prowler-cloud/prowler/pull/12813)
### 🔐 Security
- `libsqlite3-0`, `gzip`, `perl-base`, `libssh2-1t64` and `libpcre2-8-0` upgraded in the SDK container image, patching nine high Debian CVEs [(#12804)](https://github.com/prowler-cloud/prowler/pull/12804)
- PowerShell from 7.5.9 to 7.5.11 in the SDK container image, bundling .NET runtime 9.0.20 and patching CVE-2026-62901 [(#12811)](https://github.com/prowler-cloud/prowler/pull/12811)
- Bumped `anyio` to 4.14.2 to resolve CVE-2026-63374 [(#12848)](https://github.com/prowler-cloud/prowler/pull/12848)
---
## [5.42.0] (Prowler v5.42.0)
### 🚀 Added
@@ -0,0 +1 @@
`inspector2_coverage_scan_status_active`, `inspector2_coverage_recently_scanned`, `inspector2_active_findings_no_known_exploited_vulnerabilities`, `inspector2_active_findings_kev_within_due_date`, `inspector2_active_findings_within_max_age`, `elbv2_listener_fips_tls_enabled` and `transfer_server_fips_security_policy_enabled` checks for AWS provider, covering FedRAMP 20x Class C vulnerability detection, CISA KEV remediation and FIPS cryptography rules; the KEV checks require `inspector2:BatchGetFindingDetails`, now in the Prowler additions policy
@@ -0,0 +1 @@
Bootstrap STS calls now try up to two more regions of the partition declared in `PROWLER_AWS_PARTITION` when the first one cannot be reached, so a deployment that routes to only one region of its partition no longer fails on an endpoint it has no path to. This covers validating credentials, assuming a role and getting an MFA session token
@@ -1 +0,0 @@
`PROWLER_AWS_BOTO3_RETRIES_MAX_ATTEMPTS` environment variable to set the Boto3 retries for deployments without CLI flags
@@ -1 +0,0 @@
STS calls after role assumption use the answering region, avoiding a second wait for an unreachable partition region
@@ -0,0 +1 @@
Azure Defender security contacts and Key Vault key rotation policies now use the endpoints of the selected cloud (`--azure-region`) instead of the hardcoded `management.azure.com` and `vault.azure.net` hosts, so both work on `AzureUSGovernment` and `AzureChinaCloud`
@@ -0,0 +1 @@
`FedRAMP-20x-FRR-Class-C` universal compliance framework (`fedramp_20x_frr_class_c_2026`) with the 158 provider rules of the FedRAMP 20x Class C ruleset from the FedRAMP Consolidated Rules 2026 for AWS, Azure, GCP, Kubernetes and M365
@@ -0,0 +1 @@
`FedRAMP-20x-KSI` universal compliance framework (`fedramp_20x_ksi_2026`) with the 46 Key Security Indicators from the FedRAMP Consolidated Rules 2026 mapped for AWS, Azure, GCP, Kubernetes and M365
@@ -0,0 +1 @@
`fedramp_20x_ksi_low_aws`, `fedramp_20x_ksi_low_azure` and `fedramp_20x_ksi_low_gcp` FedRAMP 20x Phase One pilot frameworks, superseded by `fedramp_20x_ksi_2026`
@@ -0,0 +1 @@
`security_2sv_enforced` reports domain-wide 2-Step Verification failures as FAIL even when every failing setting is overridden for a group or organizational unit
@@ -0,0 +1 @@
`KeyError` in M365 Defender malware, anti-phishing and inbound anti-spam checks when the tenant has Standard or Strict preset security policies
@@ -1 +0,0 @@
OCI API key credentials accept a home region to bootstrap identity calls, so tenancies not subscribed to us-ashburn-1 can connect
@@ -1 +0,0 @@
`pyjwt` from 2.13.0 to 2.14.0, patching GHSA-ffc3-869f-jxw9
@@ -0,0 +1 @@
Bumped `anyio` to 4.14.2 to resolve CVE-2026-63374
@@ -0,0 +1 @@
`libsqlite3-0`, `gzip`, `perl-base`, `libssh2-1t64` and `libpcre2-8-0` upgraded in the SDK container image, patching nine high Debian CVEs
@@ -0,0 +1 @@
PowerShell from 7.5.9 to 7.5.11 in the SDK container image, bundling .NET runtime 9.0.20 and patching CVE-2026-62901
@@ -0,0 +1 @@
`smn_topic_subscriptions` check for Huawei Cloud provider: SMN topics have at least one subscription configured
@@ -1 +0,0 @@
Pass the E2E AWS credentials to the UI E2E workflow through environment variables instead of template expansion
+1 -1
View File
@@ -52,7 +52,7 @@ class _MutableTimestamp:
timestamp = _MutableTimestamp(datetime.today())
timestamp_utc = _MutableTimestamp(datetime.now(timezone.utc))
prowler_version = "5.44.1"
prowler_version = "5.43.0"
html_logo_url = "https://github.com/prowler-cloud/prowler/"
square_logo_img = "https://raw.githubusercontent.com/prowler-cloud/prowler/dc7d2d5aeb92fdf12e8604f42ef6472cd3e8e889/docs/img/prowler-logo-black.png"
aws_logo = "https://user-images.githubusercontent.com/38561120/235953920-3e3fba08-0795-41dc-b480-9bea57db9f2e.png"
+6 -13
View File
@@ -112,7 +112,7 @@ class AwsProvider(Provider):
def __init__(
self,
retries_max_attempts: Optional[int] = None,
retries_max_attempts: int = 3,
role_arn: str = None,
session_duration: int = 3600,
external_id: str = None,
@@ -141,7 +141,6 @@ class AwsProvider(Provider):
Args:
- retries_max_attempts: The maximum number of retries for the AWS client.
Defaults to the PROWLER_AWS_BOTO3_RETRIES_MAX_ATTEMPTS environment variable or, if unset, to 3.
- role_arn: The ARN of the IAM role to assume.
- session_duration: The duration of the session in seconds, between 900 and 43200.
- external_id: The external ID to use when assuming the IAM role.
@@ -1231,8 +1230,7 @@ class AwsProvider(Provider):
Args:
- session: The AWS session object
- assumed_role_info: The AWSAssumeRoleInfo object. Its sts_region is
updated to the region that answered, so later calls go straight there
- assumed_role_info: The AWSAssumeRoleInfo object
Returns:
- AWSCredentials: The AWS credentials for the assumed role
@@ -1258,14 +1256,11 @@ class AwsProvider(Provider):
mfa_info = AwsProvider.input_role_mfa_token_and_code()
assume_role_arguments["SerialNumber"] = mfa_info.arn
assume_role_arguments["TokenCode"] = mfa_info.totp
sts_region, assumed_credentials = (
AwsProvider.sts_call_with_partition_failover(
session,
assumed_role_info.sts_region,
lambda sts_client: sts_client.assume_role(**assume_role_arguments),
)
_, assumed_credentials = AwsProvider.sts_call_with_partition_failover(
session,
assumed_role_info.sts_region,
lambda sts_client: sts_client.assume_role(**assume_role_arguments),
)
assumed_role_info.sts_region = sts_region
# Convert the UTC datetime object to your local timezone
credentials_expiration_local_time = (
assumed_credentials["Credentials"]["Expiration"]
@@ -1563,8 +1558,6 @@ class AwsProvider(Provider):
session,
assumed_role_information,
)
# Validate where the role was assumed, not where it timed out
aws_region = assumed_role_information.sts_region
session = Session(
aws_access_key_id=assumed_role_credentials.aws_access_key_id,
aws_secret_access_key=assumed_role_credentials.aws_secret_access_key,
+54 -157
View File
@@ -122,21 +122,6 @@
"aws-us-gov": []
}
},
"account-access": {
"regions": {
"aws": [
"ap-southeast-1",
"us-west-1"
],
"aws-cn": [],
"aws-eusc": [],
"aws-iso": [],
"aws-iso-b": [],
"aws-iso-e": [],
"aws-iso-f": [],
"aws-us-gov": []
}
},
"acm": {
"regions": {
"aws": [
@@ -257,24 +242,6 @@
]
}
},
"agent-registry": {
"regions": {
"aws": [
"ap-northeast-1",
"ap-southeast-2",
"eu-west-1",
"us-east-1",
"us-west-2"
],
"aws-cn": [],
"aws-eusc": [],
"aws-iso": [],
"aws-iso-b": [],
"aws-iso-e": [],
"aws-iso-f": [],
"aws-us-gov": []
}
},
"ahl": {
"regions": {
"aws": [
@@ -299,20 +266,14 @@
"aiops": {
"regions": {
"aws": [
"ap-east-1",
"ap-northeast-1",
"ap-south-1",
"ap-southeast-1",
"ap-southeast-2",
"ap-southeast-5",
"ap-southeast-7",
"eu-central-1",
"eu-north-1",
"eu-south-2",
"eu-west-1",
"us-east-1",
"us-east-2",
"us-west-2"
"eu-south-2"
],
"aws-cn": [],
"aws-eusc": [],
@@ -1012,15 +973,12 @@
"aws": [
"ap-northeast-1",
"ap-northeast-2",
"ap-northeast-3",
"ap-south-1",
"ap-southeast-1",
"ap-southeast-2",
"ap-southeast-5",
"ca-central-1",
"ca-west-1",
"eu-central-1",
"eu-central-2",
"eu-south-1",
"eu-south-2",
"eu-west-1",
@@ -1587,11 +1545,37 @@
]
}
},
"awstransform": {
"regions": {
"aws": [
"ap-northeast-1",
"ap-northeast-2",
"ap-south-1",
"ap-southeast-2",
"ca-central-1",
"eu-central-1",
"eu-west-2",
"sa-east-1",
"us-east-1"
],
"aws-cn": [],
"aws-eusc": [],
"aws-iso": [],
"aws-iso-b": [],
"aws-iso-e": [],
"aws-iso-f": [],
"aws-us-gov": []
}
},
"b2bi": {
"regions": {
"aws": [
"ap-south-2",
"ap-southeast-2",
"ca-central-1",
"eu-central-1",
"eu-west-1",
"eu-west-3",
"us-east-1",
"us-east-2",
"us-west-2"
@@ -1905,7 +1889,6 @@
"ap-northeast-1",
"ap-northeast-2",
"ap-south-1",
"ap-south-2",
"ap-southeast-1",
"ap-southeast-2",
"ap-southeast-5",
@@ -1921,7 +1904,6 @@
"sa-east-1",
"us-east-1",
"us-east-2",
"us-west-1",
"us-west-2"
],
"aws-cn": [],
@@ -4220,33 +4202,16 @@
"dax": {
"regions": {
"aws": [
"af-south-1",
"ap-east-1",
"ap-east-2",
"ap-northeast-1",
"ap-northeast-2",
"ap-northeast-3",
"ap-south-1",
"ap-south-2",
"ap-southeast-1",
"ap-southeast-2",
"ap-southeast-3",
"ap-southeast-4",
"ap-southeast-5",
"ap-southeast-6",
"ap-southeast-7",
"ca-central-1",
"ca-west-1",
"eu-central-1",
"eu-central-2",
"eu-north-1",
"eu-south-1",
"eu-south-2",
"eu-west-1",
"eu-west-2",
"eu-west-3",
"il-central-1",
"mx-central-1",
"sa-east-1",
"us-east-1",
"us-east-2",
@@ -6041,10 +6006,8 @@
"evs": {
"regions": {
"aws": [
"ap-east-2",
"ap-northeast-1",
"ap-northeast-2",
"ap-northeast-3",
"ap-south-1",
"ap-south-2",
"ap-southeast-1",
@@ -6056,11 +6019,9 @@
"eu-central-2",
"eu-north-1",
"eu-south-1",
"eu-south-2",
"eu-west-1",
"eu-west-2",
"eu-west-3",
"il-central-1",
"mx-central-1",
"sa-east-1",
"us-east-1",
@@ -9257,12 +9218,9 @@
"aws": [
"af-south-1",
"ap-east-1",
"ap-east-2",
"ap-northeast-1",
"ap-northeast-2",
"ap-northeast-3",
"ap-south-1",
"ap-south-2",
"ap-southeast-1",
"ap-southeast-2",
"ap-southeast-5",
@@ -9270,17 +9228,12 @@
"ca-central-1",
"ca-west-1",
"eu-central-1",
"eu-central-2",
"eu-north-1",
"eu-south-1",
"eu-south-2",
"eu-west-1",
"eu-west-2",
"eu-west-3",
"il-central-1",
"me-central-1",
"me-south-1",
"mx-central-1",
"sa-east-1",
"us-east-1",
"us-east-2",
@@ -10012,9 +9965,7 @@
"aws-iso-b": [],
"aws-iso-e": [],
"aws-iso-f": [],
"aws-us-gov": [
"us-gov-west-1"
]
"aws-us-gov": []
}
},
"neptune": {
@@ -10476,9 +10427,7 @@
"us-west-2"
],
"aws-cn": [],
"aws-eusc": [
"eusc-de-east-1"
],
"aws-eusc": [],
"aws-iso": [],
"aws-iso-b": [],
"aws-iso-e": [],
@@ -10953,10 +10902,7 @@
"aws-iso-b": [],
"aws-iso-e": [],
"aws-iso-f": [],
"aws-us-gov": [
"us-gov-east-1",
"us-gov-west-1"
]
"aws-us-gov": []
}
},
"pca-connector-scep": {
@@ -11018,8 +10964,6 @@
"ap-southeast-1",
"ap-southeast-2",
"ap-southeast-3",
"ap-southeast-4",
"ap-southeast-6",
"eu-central-1",
"eu-north-1",
"eu-south-1",
@@ -13470,21 +13414,6 @@
"aws-us-gov": []
}
},
"securityagent": {
"regions": {
"aws": [
"us-east-1",
"us-west-2"
],
"aws-cn": [],
"aws-eusc": [],
"aws-iso": [],
"aws-iso-b": [],
"aws-iso-e": [],
"aws-iso-f": [],
"aws-us-gov": []
}
},
"securityhub": {
"regions": {
"aws": [
@@ -13948,7 +13877,6 @@
"ap-southeast-3",
"ap-southeast-4",
"ap-southeast-5",
"ap-southeast-6",
"ap-southeast-7",
"ca-central-1",
"ca-west-1",
@@ -14911,18 +14839,6 @@
"aws-us-gov": []
}
},
"supportauthz": {
"regions": {
"aws": [],
"aws-cn": [],
"aws-eusc": [],
"aws-iso": [],
"aws-iso-b": [],
"aws-iso-e": [],
"aws-iso-f": [],
"aws-us-gov": []
}
},
"sustainability": {
"regions": {
"aws": [
@@ -15146,10 +15062,8 @@
"aws": [
"ap-northeast-1",
"ap-south-1",
"ap-southeast-1",
"ap-southeast-2",
"eu-central-1",
"eu-north-1",
"eu-west-1",
"us-east-1",
"us-east-2",
@@ -15169,28 +15083,20 @@
"timestream-influxdb": {
"regions": {
"aws": [
"af-south-1",
"ap-east-1",
"ap-northeast-1",
"ap-northeast-2",
"ap-northeast-3",
"ap-south-1",
"ap-south-2",
"ap-southeast-1",
"ap-southeast-2",
"ap-southeast-3",
"ap-southeast-4",
"ap-southeast-7",
"ca-central-1",
"eu-central-1",
"eu-central-2",
"eu-north-1",
"eu-south-1",
"eu-south-2",
"eu-west-1",
"eu-west-2",
"eu-west-3",
"il-central-1",
"me-central-1",
"mx-central-1",
"sa-east-1",
@@ -15256,6 +15162,29 @@
]
}
},
"tnb": {
"regions": {
"aws": [
"ap-northeast-2",
"ap-southeast-2",
"ca-central-1",
"eu-central-1",
"eu-north-1",
"eu-south-2",
"eu-west-3",
"sa-east-1",
"us-east-1",
"us-west-2"
],
"aws-cn": [],
"aws-eusc": [],
"aws-iso": [],
"aws-iso-b": [],
"aws-iso-e": [],
"aws-iso-f": [],
"aws-us-gov": []
}
},
"transcribe": {
"regions": {
"aws": [
@@ -15508,38 +15437,7 @@
"uxc": {
"regions": {
"aws": [
"af-south-1",
"ap-east-1",
"ap-east-2",
"ap-northeast-1",
"ap-northeast-2",
"ap-northeast-3",
"ap-south-1",
"ap-south-2",
"ap-southeast-1",
"ap-southeast-2",
"ap-southeast-3",
"ap-southeast-4",
"ap-southeast-5",
"ap-southeast-6",
"ap-southeast-7",
"ca-central-1",
"ca-west-1",
"eu-central-1",
"eu-central-2",
"eu-north-1",
"eu-south-1",
"eu-south-2",
"eu-west-1",
"eu-west-2",
"eu-west-3",
"il-central-1",
"mx-central-1",
"sa-east-1",
"us-east-1",
"us-east-2",
"us-west-1",
"us-west-2"
"us-east-1"
],
"aws-cn": [],
"aws-eusc": [],
@@ -16225,7 +16123,6 @@
"aws": [
"ap-east-1",
"ap-northeast-2",
"ap-southeast-1",
"ap-southeast-5",
"eu-south-2",
"me-central-1",
+2 -23
View File
@@ -2,10 +2,7 @@ import os
from botocore.config import Config
from prowler.providers.aws.exceptions.exceptions import (
AWSInvalidBoto3RetriesError,
AWSInvalidBoto3TimeoutError,
)
from prowler.providers.aws.exceptions.exceptions import AWSInvalidBoto3TimeoutError
AWS_STS_GLOBAL_ENDPOINT_REGION = "us-east-1"
AWS_REGION_US_EAST_1 = "us-east-1"
@@ -30,28 +27,10 @@ def get_boto3_timeout_from_env(name: str, default: int) -> int:
return int(raw)
def get_boto3_retries_from_env(name: str, default: int) -> int:
"""Non-negative integer retries read from the environment, or default when unset."""
raw = os.getenv(name, "").strip()
if not raw:
return default
if not raw.isdecimal():
raise AWSInvalidBoto3RetriesError(
file=os.path.basename(__file__),
message=f"{name} must be a non-negative integer number of retries, got {raw!r}",
)
return int(raw)
def get_default_session_config() -> Config:
return Config(
user_agent_extra=BOTO3_USER_AGENT_EXTRA,
retries={
"max_attempts": get_boto3_retries_from_env(
"PROWLER_AWS_BOTO3_RETRIES_MAX_ATTEMPTS", BOTO3_RETRIES_MAX_ATTEMPTS
),
"mode": "standard",
},
retries={"max_attempts": BOTO3_RETRIES_MAX_ATTEMPTS, "mode": "standard"},
connect_timeout=get_boto3_timeout_from_env(
"PROWLER_AWS_BOTO3_CONNECT_TIMEOUT", BOTO3_CONNECT_TIMEOUT
),
@@ -82,10 +82,6 @@ class AWSBaseException(ProwlerException):
"message": "The Boto3 timeout configured through the environment is invalid",
"remediation": "Set PROWLER_AWS_BOTO3_CONNECT_TIMEOUT and PROWLER_AWS_BOTO3_READ_TIMEOUT to a positive integer number of seconds.",
},
(1919, "AWSInvalidBoto3RetriesError"): {
"message": "The Boto3 retries configured through the environment are invalid",
"remediation": "Set PROWLER_AWS_BOTO3_RETRIES_MAX_ATTEMPTS to a non-negative integer, 0 disables retries.",
},
}
def __init__(self, code, file=None, original_exception=None, message=None):
@@ -248,12 +244,3 @@ class AWSInvalidBoto3TimeoutError(AWSBaseException):
super().__init__(
1918, file=file, original_exception=original_exception, message=message
)
class AWSInvalidBoto3RetriesError(AWSBaseException):
"""Boto3 retries configured through the environment are not a non-negative integer."""
def __init__(self, file=None, original_exception=None, message=None):
super().__init__(
1919, file=file, original_exception=original_exception, message=message
)
+1 -1
View File
@@ -85,7 +85,7 @@ class S3:
aws_access_key_id: str = None,
aws_secret_access_key: str = None,
aws_session_token: Optional[str] = None,
retries_max_attempts: Optional[int] = None,
retries_max_attempts: int = 3,
regions: set = set(),
) -> None:
"""

Some files were not shown because too many files have changed in this diff Show More