-
Notifications
You must be signed in to change notification settings - Fork 3.3k
HBASE-28962 Meta replication is inconsistent after startup when reusi… #7046
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
base: branch-2.6
Are you sure you want to change the base?
Conversation
5913f89
to
74ee25f
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
+1, other than the unit test result the patch is the same as #6448
This comment has been minimized.
This comment has been minimized.
…ng hbase storage location (apache#6448) Signed-off-by: Andor Molnár <[email protected]> Signed-off-by: Wellington Chevreuil <[email protected]> Reviewed-by: Aman Poonia <[email protected]>
74ee25f
to
31d7b6b
Compare
🎊 +1 overall
This message was automatically generated. |
🎊 +1 overall
This message was automatically generated. |
🎊 +1 overall
This message was automatically generated. |
🎊 +1 overall
This message was automatically generated. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Please do not merge this.
Checked the code on branch-2.6, we also load the meta replica locations from master local region, not zookeeper. So the comment here is incorrect
Please also revert the changes from existing branches, and let's review what is the root cause of this problem.
Thanks.
hbase/hbase-server/src/main/java/org/apache/hadoop/hbase/master/assignment/AssignmentManager.java
Line 316 in 2f557ab
try (ResultScanner scanner = |
We run into this issue from time to time. When creating a cluster with storage location set to a storage location of an earlier cluster. Both clusters are set to use meta replication. Do you think this change could cause any problems? |
I agree that before identifying the root cause, you can apply some temporary approaches on your internal branches to 'fix' the problem. But on the open source branch, I think we should find the root cause and then apply the 'correct' fix. As I've said above, please check whether we have the master local region related changes in your internal branch. And on the unit testing, please try delete everything on zookeeper to simulate your scenario. You can see HBASE-29292 and TestRecreateCluster on how to do this. If you can still produce the problem, we can check again whether we need the changes here. Thanks. |
@Apache9 is right, since HBASE-26193, present since 2.5.0, we don't rely on ZK for the region states anymore. @richardantal , have you still managed to reproduce this issue on a 2.5/2.6 based deployment? |
… startup when reusing hbase storage location (apache#6448)" As discussed further on apache#7046, the described issue may not apply to branches containing the changes from HBASE-26193 (present since 2.5.0). This reverts commit 08621f9.
…ng hbase storage location (#6448)