[CELEBORN-1651] Support ratio threshold of unhealthy disks for excluding worker #2812
+21
−6
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
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Outdated suggestions cannot be applied.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.
What changes were proposed in this pull request?
Support ratio threshold of unhealthy disks for excluding worker with
celeborn.master.excludeWorker.unhealthyDiskRatioThreshold.Why are the changes needed?
We often encounter issues such as disk input/output errors in production practice. When a bad disk occurs, the worker will be maintained to decommission for repairing the machine disk. The reason is that generally the fault will be repaired in time after it is discovered. It is possible that the machine will not trigger all disk failures if it is out of warranty. It can be replaced directly when it is under warranty. If the disk fails after it is out of warranty, you need to purchase the disk yourself for replacement. At the same time, submitting the disk for repair at one time will affect the failure rate judgment of the system group and scenario. In addition, the occurrence of bad disks will bring about some management problems, such as continuous alarms, and the handling of disk failures is relatively customized.
Therefore, it's recommended to configure ratio threshold of unhealthy disks for excluding worker, which does not need to wait for all unhealthy disks to exclude corresponding worker.
Does this PR introduce any user-facing change?
Introduce
celeborn.master.excludeWorker.unhealthyDiskRatioThresholdto configure max ratio of unhealthy disks for excluding worker.How was this patch tested?
Cluster test.