domino_admin_toolkit.checks.info.test_dmm_spark module

DMM Spark Executor Inspection (informational)

Data source: Spark worker HTTP UI (spark3-worker-0.spark3-worker-headless.<compute>, port 8081) — both the JSON endpoint (/json/) and the HTML root page, which is parsed with lxml to extract executor state.

Question answered: “What Spark executors are currently running or finished on the DMM Spark worker, and what app/user owns each?”

What this check does NOT cover:

• Whether the spark3-worker pod is present and healthy — see test_dmm_pods_list.

• Whether DMM jobs are stuck — see test_dmm_ingestion_jobs_status.

• Whether the Redis queues are backing up — see test_dmm_redis_queues.

This check remains info-only: it queries an HTTP UI rather than a structured data source, and its output is best read alongside the asserting DMM checks to understand whether a stuck-in-processing ingestion job is due to Spark executor starvation.

domino_admin_toolkit.checks.info.test_dmm_spark.job_states(is_running, html_scrape)

Description:

Returns the spark job states for either running or finished jobs

domino_admin_toolkit.checks.info.test_dmm_spark.print_worker_executor_jobs(worker_data, html_scrape, is_running)

Description:

Logs executor data for running and finished jobs

domino_admin_toolkit.checks.info.test_dmm_spark.test_dmm_spark_finished_executors(k8s_client, spark3_worker_exists)

Description:

Print executor jobs that have stopped running (successfully or not)

Result:

Print data about finished spark jobs

domino_admin_toolkit.checks.info.test_dmm_spark.test_dmm_spark_running_executors(k8s_client, spark3_worker_exists)

Description:

Print executor jobs that are either currently or pending running

Result:

Print spark running jobs