Trevor Hickey
Trevor Hickey
I know this change may not be exactly what you want for pipeline behavior. For us, its sort of a stop gap to improve cluster stability. Since redis going away...
Then we will need to ensure that the pipeline properly shuts down when a stage fails and that the worker also shuts down when a pipeline fails. Note to self,...
based on the color of the background, I think it loaded correctly, its just that a phone's viewport is too small to see any of the GUI elements
@80degreeswest close?
> It looks like you're trying to create a capture interface with the least involvement in changes on the Redis side - I'd prefer fixes to the usage to make...
> > It looks like you're trying to create a capture interface with the least involvement in changes on the Redis side - I'd prefer fixes to the usage to...
@werkt ready for re-review
This PR is similar in spirit to: https://github.com/bazelbuild/bazel-buildfarm/pull/1040 I agree that in both cases, @jerrymarino and I are taping over the problems to keep the workers up an running. It's...
It's likely from here because the operation expired from redis somehow: https://github.com/bazelbuild/bazel-buildfarm/blob/74f7799d08c0b6bd02a8730d873338f461a2c6be/src/main/java/build/buildfarm/instance/shard/ShardInstance.java#L2502-L2505
Good job profiling this. It's nice to see you got a performance boost as well. My only suggestion is to make the UUID fetching something like a [Supplier](https://docs.oracle.com/javase/8/docs/api/java/util/function/Supplier.html). That way,...