Skip to content

agent: expose DISABLE_IDLE_TASKS and DISABLE_FAILING_TASKS separately#2802

Merged
jshearer merged 2 commits intomasterfrom
agent/independently_control_disablement
Apr 2, 2026
Merged

agent: expose DISABLE_IDLE_TASKS and DISABLE_FAILING_TASKS separately#2802
jshearer merged 2 commits intomasterfrom
agent/independently_control_disablement

Conversation

@jshearer
Copy link
Copy Markdown
Contributor

Summary

  • Replaces DISABLE_ABANDONED_TASKS with two independent env vars: DISABLE_IDLE_TASKS and DISABLE_FAILING_TASKS, both defaulting to false
  • Allows enabling auto-disable for idle tasks (30d no data movement) without also auto-disabling chronically failing tasks, and vice versa

Context

Discussion about abandoned task tracking surfaced that many chronically failing tasks are still actively moving data, and the upshot is that we should be able to enable idle-task disablement independently while we work on improving shard failure alerting before acting on chronically failing tasks.

@jshearer jshearer force-pushed the agent/independently_control_disablement branch from b3bcf33 to 107bf68 Compare March 23, 2026 20:41
@jshearer jshearer requested a review from a team March 23, 2026 20:42
@jshearer jshearer force-pushed the agent/independently_control_disablement branch from 107bf68 to 1cabb8d Compare March 24, 2026 13:45
The single `DISABLE_ABANDONED_TASKS` env var prevented turning on auto-disable for idle tasks without also enabling it for chronically failing tasks. Replaces it with `DISABLE_IDLE_TASKS` and `DISABLE_FAILING_TASKS` so each can be enabled independently.
@jshearer jshearer force-pushed the agent/independently_control_disablement branch from 1cabb8d to 1d51882 Compare March 24, 2026 21:52
@jshearer jshearer self-assigned this Mar 25, 2026
@jshearer jshearer requested a review from GregorShear March 30, 2026 16:52
Copy link
Copy Markdown
Contributor

@GregorShear GregorShear Mar 31, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

the concept of "abandoned tasks" isn't going away, right? DISABLE_ABANDONED_TASKS is splitting into two more specific flags, but we'll still use the term "abandoned" generally? (mostly just trying to get up to speed here...)

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yeah that's right. I tried to think of naming for these env vars that contained the word ABANDONED but the ones I came up with were too wordy imo so I went with what I have here

Copy link
Copy Markdown
Contributor

@GregorShear GregorShear left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Consider testing isolation between the two flags, i.e. a chronically failing task is not disabled when the idle flag is on. Not a blocker...

…NG_TASKS`

* A chronically failing task with an expired grace period remains enabled when only `disable_idle_tasks` is set
* An idle task with an expired grace period remains enabled when only `disable_failing_tasks` is set
@jshearer jshearer merged commit 7011478 into master Apr 2, 2026
13 of 14 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants