-
Notifications
You must be signed in to change notification settings - Fork 948
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
Processing web console line queue takes long time to stop in self hosted runner #530
Comments
@jdsdc those log means the runner didn't receive any output from the action. can you try to enable step debug to see whether the action will print out more outputs to tells you where is the time spent? |
@TingluoHuang , thanks for the quick answer. I have a clue where the time is spent. When i look in the log it seems like the cache is not on the local machine. Does the cache action save to this url: https://artifactcache.actions.githubusercontent.com/ If so that might explain the time delay?
|
In general the cache action is really meant for the hosted runners that are new for each invocation. It runs in Azure along with the rest of our runner infrastructure so it’s network close. If you have your own self-hosted machines that are largely static I would expect the normal tool caches to still be present between runs. |
@chrispat, i see what you mean. I should have some sort of local storage that i manage myself, fileserver or some local docker repository to store images. With normal tool chains i guess you mean gradle caches and the likes. Thanks. Think this can be closed now and conclude that the action cache is not the best solution for self hosted runners. |
Running a workflow with action actions/cache@v2 takes a long time to execute when job is run in self hosted mode. Works fine when running on github servers. From what i can see in the log the action itself seems to work fine but takes several minutes to finish cause of wait for a web console line queue. When running on github it takes about 1 minute.
The self hosted runner environment:
OS: Ubuntu 20.04
Runner version: 2.263.0
As you can see from this log line an action is taken to stop a queue.
[2020-06-10 14:46:37Z INFO JobServerQueue] Stop aggressive process web console line queue.
And the nex log line:
[2020-06-10 14:50:32Z INFO JobServerQueue] Try to append 1 batches web console lines for record 'a04c83ac-b35b-5e6c-22c8-b60f8cc735cb', success rate: 1/1.
And the time to finish for the action it takes about 4 minutes. And i don't know why there is such a long wait time for this web console line queue.
Extract from the log:
This is the relevant action part from the workflow file;
Is there some configuration i need to change when running with a self-hosted runner? It seems to me that there is som lag in the communication between the self hosted runner and github.
The text was updated successfully, but these errors were encountered: