Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

ExecutionTimeout to fail executions instead of jobs #3974

Merged
merged 12 commits into from May 13, 2024
Merged

ExecutionTimeout to fail executions instead of jobs #3974

merged 12 commits into from May 13, 2024

Conversation

wdbaruni
Copy link
Collaborator

@wdbaruni wdbaruni commented May 6, 2024

This work is a pre-requirement for job queues as the orchestrator is currently failing jobs that take longer than ExecutionTimeout from the time they were submitted. ExecutionTimeout should only fail executions that have been running for too long, and not fail jobs. If a job haven't started and has been in the queue for too long, then this timeout should take effect as there will be other config to timeout of fail a job has been in the queue for too long or has been retried too many times.

A side effect of this PR is that timed out executions will be retried by the scheduler on other nodes until one succeed or no more nodes to retry. This will be handled in next job queue related PRs

Copy link

coderabbitai bot commented May 6, 2024

Important

Auto Review Skipped

Auto reviews are disabled on this repository.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.


Tips

Chat

There are 3 ways to chat with CodeRabbit:

  • Review comments: Directly reply to a review comment made by CodeRabbit. Example:
    • I pushed a fix in commit <commit_id>.
    • Generate unit testing code for this file.
    • Open a follow-up GitHub issue for this discussion.
  • Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
    • @coderabbitai generate unit testing code for this file.
    • @coderabbitai modularize this function.
  • PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
    • @coderabbitai generate interesting stats about this repository and render them as a table.
    • @coderabbitai show all the console.log statements in this repository.
    • @coderabbitai read src/utils.ts and generate unit testing code.
    • @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (invoked as PR comments)

  • @coderabbitai pause to pause the reviews on a PR.
  • @coderabbitai resume to resume the paused reviews.
  • @coderabbitai review to trigger a review. This is useful when automatic reviews are disabled for the repository.
  • @coderabbitai resolve resolve all the CodeRabbit review comments.
  • @coderabbitai help to get help.

Additionally, you can add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.

CodeRabbit Configration File (.coderabbit.yaml)

  • You can programmatically configure CodeRabbit by adding a .coderabbit.yaml file to the root of your repository.
  • Please see the configuration documentation for more information.
  • If your editor has YAML language server enabled, you can add the path at the top of this file to enable auto-completion and validation: # yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json

Documentation and Community

  • Visit our Documentation for detailed information on how to use CodeRabbit.
  • Join our Discord Community to get help, request features, and share feedback.
  • Follow us on X/Twitter for updates and announcements.

This work is a pre-requirement for job queues as the orchestrator is currently failing jobs that take longer than ExecutionTimeout from the time they were submitted. ExecutionTimeout should only fail executions that have been running for too long, and not fail jobs. If a job haven't started and has been in the queue for too long, then this timeout should take effect as there will be other config to timeout of fail a job has been in the queue for too long or has been retried too many times.

A side effect of this PR is that timed out executions will be retried by the scheduler on other nodes until one succeed or no more nodes to retry. This will be handled in next job queue related PRs
Copy link
Member

@frrist frrist left a comment

Choose a reason for hiding this comment

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

Three major points of concern:

  • Parallelism in Housekeeping has several issues, one of which the potential for panicking. Suggest adopting the workerpool package I mentioned instead of implementing our own.
  • Don't embed context.Context in the housekeeping structure.
  • We need to be mocking time in our tests, especially as this area of the code base grows.

pkg/orchestrator/housekeeping.go Outdated Show resolved Hide resolved
pkg/models/event.go Outdated Show resolved Hide resolved
pkg/models/execution.go Outdated Show resolved Hide resolved
EventTopicJobScheduling models.EventTopic = "Scheduling"
EventTopicJobSubmission models.EventTopic = "Submission"
EventTopicJobScheduling models.EventTopic = "Scheduling"
EventTopicExecutionTimeout models.EventTopic = "Exec Timeout"
Copy link
Member

Choose a reason for hiding this comment

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

I think we should avoid including spaces in topic names. Let's opt for Camel Case.

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

Topics main use is improving readability of events and let users understand which component or lifecycle or job orchestration the event is related. Not an internal detail and not intended to be consumed to other applications.

As the goal is improve human readability, allowing spaces can be improve the user experience and will improve how we display them on the CLI and UI without having to truncate them or wrap them at weird palces

pkg/orchestrator/events.go Outdated Show resolved Hide resolved
pkg/orchestrator/housekeeping.go Outdated Show resolved Hide resolved
pkg/orchestrator/housekeeping_test.go Show resolved Hide resolved
pkg/orchestrator/scheduler/batch_job_test.go Outdated Show resolved Hide resolved
sample_config.yaml Outdated Show resolved Hide resolved
pkg/orchestrator/housekeeping_test.go Outdated Show resolved Hide resolved
@wdbaruni
Copy link
Collaborator Author

Parallelism in Housekeeping has several issues, one of which the potential for panicking. Suggest adopting the workerpool package I mentioned instead of implementing our own.

Not sure about the several issues. You pointed to one issue with a wrong call to waitGroup.Done() and that was fixed. You forgot to mention the workerpool package, but the semaphore seem to be working fine. The other requirement I have is I don't want the housekeeper to start another iteration before all tasks are completed, and thats why I am using the waitGroup

Don't embed context.Context in the housekeeping structure.

Fixed

We need to be mocking time in our tests, especially as this area of the code base grows.

Answered inline

@frrist
Copy link
Member

frrist commented May 10, 2024

Here's the worker pool package: https://github.com/gammazero/workerpool

The several issues were embedded context in structure, and the potential for panic with the wait group. The look to be addressed now will give another review in a sec.

@wdbaruni wdbaruni added the comp/scheduler Related to job scheduling components label May 13, 2024
@wdbaruni
Copy link
Collaborator Author

wdbaruni commented May 13, 2024

I've previously looked into the worker pool library you mentioned, and it doesn't enable waiting for housekeeping tasks before calling the iteration as complete and starting another one. Our requirements are pretty simple and I don't see a need for such a library at this point. The semaphore is limiting the number of concurrent tasks, which is what the workerpool library does, and the waitGroup is preventing iterations from overlapping with each other, which the workerpool library is missing.

@wdbaruni wdbaruni merged commit 30b4213 into main May 13, 2024
12 checks passed
@wdbaruni wdbaruni deleted the queue branch May 13, 2024 15:28
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
comp/scheduler Related to job scheduling components
Projects
Status: Done
Development

Successfully merging this pull request may close these issues.

None yet

2 participants