Bug: Pipeline deletion via REST API does not work as expected

[ChristopherFoerster]

We have several automations for project management using REST API.

Using API Swagger UI or Python or UI: runs into 502 "bad gateway" for deletion of a pipeline.

After tests this impacts "some" models. After trying to use the API or UI to delete it the backend crashes and restarts. Means that might be more critical that expected. We are using
Frontend version: v5.1.1
Backend version: 5.1.1

PS: Other deletions of projects etc work well. Do you have any idea?

[ChristopherFoerster]

It seems to work for other models but not for this specific one. What could be reasons? How can we dubug this?

[ChristopherFoerster]

Here is the summary from https://capella-test.platform.aws.gts/api/v1/pipelines
from my perspective everything looks "OK". Anyhow we need a way to manually remove a "undeletable" pipeline job. Otherwise we are blocked in managing the model (remove model , change T4C, ....) as all depends on the pipeline.
[ { "id": 8, "t4c_model": { "id": 14, "name": "In-Flight Entertainment System", "repository": { "name": "urs-training", "id": 22, "instance": { "id": 1, "name": "6.1.0", "version": { "name": "6.1.0", "id": 6 }, "is_archived": false } } }, "git_model": { "path": "https://infinity.platform.aws.gts/gts-e2-services/capella-models/urs-training.git", "entrypoint": "In-Flight Entertainment System.aird", "revision": "master", "username": "capella-gitlab-token", "password": true, "id": 20, "primary": true, "repository_id": "158" }, "run_nightly": true, "next_run": null, "model": { "id": 29, "slug": "urs-training", "name": "urs-training", "project": { "id": 30, "name": "urs-training", "slug": "urs-training", "visibility": "internal", "type": "general" } } } ]

Backend log during deletion call:
DEBUG client="10.187.244.239:0" user="Christopher FOERSTER" trace_id="V0K5HMVCO4" method="DELETE" __init__.py:135 │ path="/api/v1/projects/urs-training/models/urs-training/backups/pipelines/8" │ query_params="force=true" message="request started" WARNING Job for pipeline 8 not found during deletion: 'No job by the id of pipeline-8 was found' interface.py:79 DEBUG client="100.64.106.242:42282" user="anonymous" trace_id="ZLY60EC5CW" method="GET" __init__.py:135 path="/metrics" query_params="" message="request started" DEBUG client="100.64.106.242:42282" user="anonymous" trace_id="ZLY60EC5CW" method="GET" __init__.py:138 path="/metrics" query_params="" status_code="200" duration="0.7934" message="request finished" DEBUG client="10.187.244.239:0" user="Christopher FOERSTER" trace_id="CAFDV9HY38" method="GET" __init__.py:135 path="/api/v1/configurations/unified" query_params="" message="request started" Stream closed EOF for capella/capella-backend-78f986dfc-s5xn6 (capella-backend)

[MoritzWeber0]

That's interesting indeed. What's the reason for the backend restart? Can you share the output of the following command for the backend pod after the restart?

k describe pods {pod_name}

How many runs has the pipeline?

[ChristopherFoerster]

@MoritzWeber0 Oana will check the logs - meanwhile the model has one backup pipeline and the pipeline was executed nightly but stopped after we tried to delete is the first time. Since then no execution automatically. Manual execution is still working even if the backup itself is failing.

(java:88): dbind-WARNING **: 07:08:21.793: AT-SPI: Error retrieving accessibility bus address: org.freedesktop.DBus.Error.ServiceUnknown: The name org.a11y.Bus was not provided by any .service files