HBASE-21070 Fix SnapshotFileCache for HBase backed by S3 #209
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
|
💔 -1 overall
This message was automatically generated. |
Apache9
requested changes
May 1, 2019
hbase-server/src/main/java/org/apache/hadoop/hbase/master/snapshot/SnapshotFileCache.java
Outdated
Show resolved
Hide resolved
hbase-server/src/main/java/org/apache/hadoop/hbase/master/snapshot/SnapshotFileCache.java
Outdated
Show resolved
Hide resolved
hbase-server/src/main/java/org/apache/hadoop/hbase/master/snapshot/SnapshotFileCache.java
Outdated
Show resolved
Hide resolved
hbase-server/src/main/java/org/apache/hadoop/hbase/master/snapshot/SnapshotFileCache.java
Outdated
Show resolved
Hide resolved
hbase-server/src/main/java/org/apache/hadoop/hbase/master/snapshot/SnapshotFileCache.java
Outdated
Show resolved
Hide resolved
hbase-server/src/main/java/org/apache/hadoop/hbase/master/snapshot/SnapshotFileCache.java
Outdated
Show resolved
Hide resolved
|
Updated. |
|
💔 -1 overall
This message was automatically generated. |
|
A simple question here, does S3 FileSystem support atomic rename? If not, what if we list and read the snapshot directories during a rename processing? |
|
@Apache9 No S3 Filesystems don't support a true atomic rename (at least none that I know of :) ). I think I have changed my mind and prefer removing the modification time since I think the performance gain is marginal. I think I will remove SnapshotFileCache changes, but leave the test so we don't regress. |
SnapshotFileCache depends on getting the last modified time of the snapshot directory, however, S3 FileSystem's do not update the last modified time of the top 'folder' when objects are added/removed. This commit adds a test for the previously fixed SnapshotFileCache.
|
Updated to only include the test. Verified that it fails without this patch and passes with it. |
|
💔 -1 overall
This message was automatically generated. |
|
@z-york If S3 does not support atomic rename, then I think there are problems here... Maybe we could see an empty directory when refreshing cache? |
|
Anyway, I think the test here is good. Maybe we can commit the test first, and open new issue to address the atomic rename related problems. In general, I think we'd better not rely on any non-trivial behavior of filesystem so we can build HBase on any possible filesystems. Of course the WAL file system is an exception. |
Apache9
approved these changes
May 4, 2019
|
@Apache9 I took a look at the code base and do not think there is an issue with atomic rename here. The issue would be: The snapshotDir exists, but the snapshot info hasn't been copied over which will fail here: https://github.com/apache/hbase/blob/master/hbase-server/src/main/java/org/apache/hadoop/hbase/snapshot/SnapshotDescriptionUtils.java#L361 which will throw. refresh cache is called by the RefreshCacheTask (which will clear on IOException), trigger refresh for testing, and getUnreferencedFiles. GetUnreferencedFiles will return no files for deletion if a corruptSnapshotException is thrown and it will be picked up the next time it is ran: https://github.com/apache/hbase/blob/master/hbase-server/src/main/java/org/apache/hadoop/hbase/master/snapshot/SnapshotHFileCleaner.java#L70 So I think it is safe. |
|
Anyways, I will merge this for now. Feel free to open a JIRA and tag me if you feel there are other cases this doesn't cover. |
asfgit
pushed a commit
that referenced
this pull request
May 6, 2019
SnapshotFileCache depends on getting the last modified time of the snapshot directory, however, S3 FileSystem's do not update the last modified time of the top 'folder' when objects are added/removed. This commit adds a test for the previously fixed SnapshotFileCache.
asfgit
pushed a commit
that referenced
this pull request
May 7, 2019
SnapshotFileCache depends on getting the last modified time of the snapshot directory, however, S3 FileSystem's do not update the last modified time of the top 'folder' when objects are added/removed. This commit adds a test for the previously fixed SnapshotFileCache.
infraio
pushed a commit
to infraio/hbase
that referenced
this pull request
Aug 17, 2020
…ache#209) SnapshotFileCache depends on getting the last modified time of the snapshot directory, however, S3 FileSystem's do not update the last modified time of the top 'folder' when objects are added/removed. This commit adds a test for the previously fixed SnapshotFileCache.
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.
SnapshotFileCache depends on getting the last modified time of the
snapshot directory, however, S3 FileSystem's do not update the
last modified time of the top 'folder' when objects are added/removed.