HBASE-29706: Alter table with reopen false should pass if coprocessors havenot changed #7575
+133
−4
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
mnpoonia
force-pushed
the
hbase_lazy_mode
branch
from
cd74717 to
4272f2c
Compare
This comment has been minimized.
This comment has been minimized.
Contributor
|
Need to run |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
...-server/src/test/java/org/apache/hadoop/hbase/master/procedure/TestModifyTableProcedure.java
Outdated
Show resolved
Hide resolved
mnpoonia
force-pushed
the
hbase_lazy_mode
branch
from
4272f2c to
3a1ebef
Compare
|
🎊 +1 overall
This message was automatically generated. |
|
🎊 +1 overall
This message was automatically generated. |
Contributor
Author
|
@apurtell @virajjasani please review when you get some free cycle. |
Apache9
reviewed
Dec 27, 2025
Contributor
Apache9
left a comment
Apache9
left a comment
There was a problem hiding this comment.
The title of the PR is different with the title of the jira...
Could someone explain more about what do we actually fix here?
Thanks.
| ) { | ||
| return false; | ||
| } | ||
| return new TreeMap<>(getProperties()).equals(new TreeMap<>(other.getProperties())); |
Contributor
There was a problem hiding this comment.
I think we could use equals directly here?
Contributor
Author
There was a problem hiding this comment.
Fixed and improved it.
|
|
||
| @Override | ||
| public int hashCode() { | ||
| return Objects.hash(className, jarPath, priority, new TreeMap<>(properties)); |
Contributor
There was a problem hiding this comment.
Better use Objects.hash to calculate hashCode for fields other than properties, and then use stream.sorted to iterator over properties and calculate hashCode on the fly.
Contributor
Author
There was a problem hiding this comment.
Done. Almost what you suggested.
| if ( | ||
| this.unmodifiedTableDescriptor.getCoprocessorDescriptors().hashCode() | ||
| != this.modifiedTableDescriptor.getCoprocessorDescriptors().hashCode() | ||
| !new HashSet<>(this.unmodifiedTableDescriptor.getCoprocessorDescriptors()) |
Contributor
There was a problem hiding this comment.
This is not on the critial read/write path so maybe this is acceptable, but a better way, is to first check size, and then, create a HashSet for one of the descriptors, and then iterator over the other one, and check whethere the HashSet contains it.
…s havenot changed
mnpoonia
force-pushed
the
hbase_lazy_mode
branch
from
3a1ebef to
d66e3ac
Compare
mnpoonia
changed the title
|
🎊 +1 overall
This message was automatically generated. |
Apache9
reviewed
Dec 29, 2025
| @Override | ||
| public int hashCode() { | ||
| int result = Objects.hash(className, jarPath, priority); | ||
| List<Map.Entry<String, String>> entries = new ArrayList<>(properties.entrySet()); |
Contributor
There was a problem hiding this comment.
Ah, checked the code, properties is already a TreeMap... So we do not need to sort here, just iterate over it is enough...
To make it safe, we can change the declaration of properties from Map to SortedMap or NavigableMap.
Contributor
Author
There was a problem hiding this comment.
Consdering this change is not in hot path should we still change the treemap to map(Navigable or sortable)?
This might become a cleanup exercise where treemap specific functionality is used elsewhere. I mean i am happy to change it but doesn't make sense to do it in this jira atleast.
Contributor
Author
There was a problem hiding this comment.
Let me know if you want it within same jira.
Contributor
There was a problem hiding this comment.
I think it is related. We want to make sure the hash code calculation is stable, so here we want to sort the entries before calculating. Since we already use TreeMap in CoprocessorDescriptorBuilder, we can make sure that the Map in CoprocessorDescriptor is a TreeMap, which is sorted, so we do not need to sort it again.
To prevent later developers break the assumption, i.e, changing the TreeMap to HashMap in CoprocessorDescriptorBuilder without any compilation errors, we'd better change the declaration in CoprocessorDescriptor to SortedMap or NavigableMap, so the constructor will not accept a HashMap any more. And we could also add some comments to explain that we must use SortedMap here as we need it for a stable hash code calculation.
Thanks.
|
🎊 +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
4 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.
No description provided.