Artificial Intelligence For Remote Cache

[bayareabear]

When remote cache is enabled, it would be good to add a field in AC entry to record how much CPU time it took to complete that AC and also a field for the total file size of it's output file. Then according to the network speed, it can calculate if it is faster to run the AC locally or just download the output from remote cache.
On the other hand, it can also set a threshhold wait time for each action accordingly, if the wait time exceeds the threshhold, then do it locally.
If the total build output is over 10GB, then soon enough the network bandwidth becomes the bottleneck.
Also, the assumption that an CAS entry referred by an AC must exist on the server makes it hard to do load balancing and horizonal scale to overcome the network bandwidth bottleneck as mentioned in this issue:
buchgr/bazel-remote#65

[buchgr]

Thanks for the idea! I believe this could be something that we'll want to implement eventually but probably only if still necessary after #7086 and #6862 have been implemented. What do you think?

[github-actions]

Thank you for contributing to the Bazel repository! This issue has been marked as stale since it has not had any activity in the last 3 years. It will be closed in the next 14 days unless any other activity occurs or one of the following labels is added: "not stale", "awaiting-bazeler". Please reach out to the triage team (@bazelbuild/triage) if you think this issue is still relevant or you are interested in getting the issue resolved.

[github-actions]

This issue has been automatically closed due to inactivity. If you're still interested in pursuing this, please reach out to the triage team (@bazelbuild/triage). Thanks!