HBASE-29682: Roll HMaster WAL in response to FlushMasterStoreRequest #7407
+7
−5
Conversation
This file contains hidden or 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
|
🎊 +1 overall
This message was automatically generated. |
|
Which WAL implementation do you use? We should be able to deal with datanode restart. The data which has been hflushed should be OK, and once there is a write error, we will try rolling the WAL and then write the data out again. |
|
🎊 +1 overall
This message was automatically generated. |
|
We're not setting The problem we see in HMasters is this: Since several DataNodes have recently restarted, they get added to the DFSClient's bad node list, and once enough DataNodes are on the bad nodes list, there are none left to write to. This makes the HMaster restart. |
Seems we have already started a log roll here? Then I do not think issuing a log roll manually can fix the problem? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.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.
I argue that in addition to flushing the master region, the
flushMasterStore()RPC should also roll the WAL files being written by the HMaster.In clusters with small numbers of DataNodes (less than 9, in my experience), a rolling restart of the DataNodes will break WAL writers, in both HMasters and RegionServers. This can be worked around on RegionServers by calling
admin.rollWALWriter()on every RegionServer after restarting each DataNode. Currently, there is no equivalent way to do this to an HMaster. I think the most elegant solution is to augmentadmin.flushMasterStore().