Make jobs run a little safer on Heroku #403
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
It turns out Heroku sends kill signals directly to all child processes of a Resque worker, which Resque is not designed to handle. That causes jobs to get killed without get logged as killed, and so they never get retried later. I'm pretty sure this is one of the causes I'm seeing of hanging import jobs for Internet Archive. See this GitHub issue for more about why the gem exists: resque/resque#1559
Since import jobs can go for a long time, it's useful to periodically persist any updates to the `Import` record so progress, warnings, and errors are visible in the API. It would be unnecessary overhead to persist every update immediately, so this opts for updating every 5 seconds (I'm thinking that if the job is shorter than that, we don't really need in-progress data anyway).
|
Alright, I’m going to go ahead and merge this, since nobody has reviewed. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
While trying to get Wayback imports working reliably, I’ve noticed that we frequently wind up with job progress stuck hanging for hours and hours and hours. Obviously we need to give up at some point on the ETL script side, but it’s not like any import jobs are actually taking that long.
Instead, what’s really happening is that Heroku has some Resque-incompatible behavior when it randomly kills dynos every so often. Specifically, it sends SIGTERM to both the Resque process and to the process’s children, but Resque’s children aren’t prepared for that — the children simply die and the parent never tracks the fact that the job didn’t finish and so never re-enqueues it to run later. The
resque-heroku-signalsgem should solve that issue.Additionally, since the Wayback import jobs are so big, it would be nice to have some sense of progress in the API and the database. This now persists updates to the
Importrecord every 5 seconds (so as to not cause a lot of overhead that slows down processing an import).