HDFS-17861. The mis-behavior of commitBlockSynchronization may cause standby namenode and observer namenode crash. #8120
+178
−0
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
…standby namenode and observer namenode crash.
|
💔 -1 overall
This message was automatically generated. |
|
💔 -1 overall
This message was automatically generated. |
Member
Author
|
All tests are passed in my local. |
Member
Author
|
@Hexiaoqiao @zhangshuyan0 @ayushtkn Have uploaded an unit test. please help review when you have free time~ Thanks a lot. |
Member
Author
|
Hi, @zhtttylz. I found other merge requests also have the same failed unit test. Could you please check it? |
|
💔 -1 overall
This message was automatically generated. |
Contributor
Thanks for the heads-up! I’ve noticed the same failure in a few other PRs as well. I’ll investigate |
Member
Author
Great! Thanks ahead. |
Contributor
|
+1. LGTM. |
zhangshuyan0
approved these changes
Jan 6, 2026
haiyang1987
reviewed
Jan 6, 2026
| dfs.mkdirs(dir); | ||
| dfs.setErasureCodingPolicy(dir, ecPolicy.getName()); | ||
| } | ||
|
|
Contributor
There was a problem hiding this comment.
please remove extra spaces.
Member
Author
There was a problem hiding this comment.
@haiyang1987 Have fixed. Thanks a lot for reviewing.
|
💔 -1 overall
This message was automatically generated. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 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Outdated suggestions cannot be applied.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.
Description of PR
Recently, our standby namenode and observer namenode crash suddenly. The error stack is as below :
There are also some other logs:
After diving into FSNamesystem#commitBlockSynchronizationsource method, we find the root cause of this critical problem is as below:
1、
deleteblock == true, So we remove a block of INodeFile successfully.2、
closeFile == true，we try to closeFileCommitBlocks.3、 In method closeFileCommitBlocks, we met an
IllegalStateException Exception, this cause edit log write failed.At this moment, the blocks metadatas in active namenode are different from standby(or observer) namenode.
4、In the next block recovery progress, all operations are success. we write an OP_CLOSE edit log.
5、Standby namenode apply that OP_CLOSE edit log and found metadata not match， then crashed.
Below are the key codes snippets:
How to test ?
Add an unit test. The result is as below:
Before applying this patch: