Add --idle-timeout CLI option to create temporary processes that exit when inactive
#1159
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
A very naive way to monitor for idle processes
Expose stats, and tweak the logic for checking idle
Put the formatting back so that it matches the rest of the codebase
While trying to get the test running, I found that the reference to the Gemfile was going one directory to high. I've corrected it here and the bin/setup command works now.
Instead of pushing the logic of idle all the way up to the CLI I thought it would be better in the GoodJob::Capsule class. I've created a `#idle?` method that does this check based on the last time a scheduler in the GoodJob::Capsule has run a job. I ran into an interesting thing that if I check the configuration object in the initialization of the GoodJob::Capsule it causes a crash because rails is not initialized yet. The configuation object can't access the rails config. I instead changed it so that in the initialize of the GoodJob::Capsule the `shutdown_on_idle_enabled` flag is set to false - meaning never shutdown on idle. Then in `#start` the GoodJob::Capsule checks the config and sets the flag so that #idle? method will work.
Modifications for shutdown on idle
I had a stale file open and overwrote the check of the event. This was after I had pushed up all the code. This also made me realize that, now one can't set less than a 10 second idle time out. The CLI will always stay up for at least 10 seconds. As mentioned in a previous issue, this may cause a support request when someone sets a idle time out of 1 second but the CLI stays up for 10 seconds then shuts down. In real life I can't imagine that someone would want to have a job server around for less than 10 seconds. Maybe this just needs to be surfaced in the documentation for the command line option. In any case, I've added code so that the wait will take the shutdown_on_idle as the time out to check the event, or it will take the default of 10 seconds. Also in my testing I realized that the `executed_at` was being set in every loop because of the `increment_empty_executions`, this meant that the time out would never happen during actual execution. I've removed the setting of the `executed_at` in this function as technically nothing executed.
Over wrote something by mistake
1 task
bensheldon
force-pushed
the
Paradem/main
branch
from
e9176cd to
c93db7e
Compare
bensheldon
force-pushed
the
Paradem/main
branch
from
c93db7e to
0ccd46a
Compare
bensheldon
force-pushed
the
Paradem/main
branch
from
0ccd46a to
249ccd5
Compare
|
Going to rename this from |
bensheldon
force-pushed
the
Paradem/main
branch
from
249ccd5 to
b14d3f7
Compare
bensheldon
changed the title
--exit-on-idle CLI option for creating temporary processes--idle-timeout CLI option to create temporary processes that exit when inactive
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.
Replaces #1146. Connects to #1140. Naming inspired by puma/puma#3209
The idea here is that, especially in a containerized environment, temporary instances of the GoodJob CLI can be spun up to execute jobs, and then they will automatically exit when no jobs remain to be executed.
Note that this feature is ignorant of the polling interval or future-scheduled jobs. So if the exit-on-idle duration is less than the polling interval and LISTEN/NOTIFY is disabled, it's possible for the process to exit even though jobs exist on the queue.