HDFS-17050. Erasure coding: fix bug for invalidating duplicated block when two ec block at the same datanode but different storage. #5753
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
… when two ec block at the same datanode but different storage.
|
💔 -1 overall
This message was automatically generated. |
|
Great catch here! I think @zhangshuyan0 should be more proper to involve to take review. |
@Hexiaoqiao Sir, thanks a lot for your reply, current modification seems to have some problems and cause the failed UT. |
|
💔 -1 overall
This message was automatically generated. |
| @@ -195,11 +195,11 @@ public void run() { | |||
|
|
|||
| if (extraSleepTime > warnThresholdMs) { | |||
| ++numGcWarnThresholdExceeded; | |||
| LOG.warn(formatMessage( | |||
| LOG.debug(formatMessage( | |||
There was a problem hiding this comment.
Why does this need to be changed to debug?
There was a problem hiding this comment.
@zhangshuyan0 Hi, shuyan, thanks for your reviewing. Here is my fault, i modify this to disable JVMPause log in my loal. I will fix it.
| if (BlockIdManager.convertToStripedID(blockInfo.getBlockId()) == blockGroupId) { | ||
| Block blockOnOldStorage = ((BlockInfoStriped) blockInfo).getBlockOnStorage(old); | ||
| if (blockOnOldStorage.getBlockId() != reportedBlock.getBlockId()) { | ||
| blockIdNotEquals = true; |
There was a problem hiding this comment.
Sorry, I'm a bit confused here. We retrieve old using index, How can blockOnOldStorage.getBlockId() and reportedBlock.getBlockId() be different?
There was a problem hiding this comment.
@zhangshuyan0 Thanks shuyan for your reviewing, this PR still has some problems, I will fix it soonly and push it. I think it is efficent for us to disscuss after then.
| final StoredReplicaState s; | ||
| if (storage.getState() == State.NORMAL) { | ||
| final DatanodeDescriptor node = storage.getDatanodeDescriptor(); | ||
| if (nodesCorrupt != null && nodesCorrupt.contains(node)) { | ||
| if (nodesCorrupt != null && nodesCorrupt.contains(node) && | ||
| (alreadyCorrupt == null || !alreadyCorrupt.contains(node))) { |
There was a problem hiding this comment.
What does this new added condition mean? I think the current implement of corruptReplicasMap can not distinguish between different internal blocks on the same datanode.
ayushtkn
reviewed
Jun 29, 2023
There was a problem hiding this comment.
the problem makes sense, I am not able to decode the solution though in one pass and I saw there is a comment saying there is some problem in the current solution, so would circle back once it is ready for review state.
but if the problem is for real, then it is indeed a nice catch :)
Comment on lines
+9
to
+11
| * <p> | ||
| * http://www.apache.org/licenses/LICENSE-2.0 | ||
| * <p> |
Sir, thanks for your replying. Yes, current solution still have some problems. I will fix it soonly and we can discuss then. |
|
Have moved to HDFS-17071. |
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.
Description of PR
Please see HDFS-17050.
How was this patch tested?
add a UT