Fix iceberg data migration test timeout #25069
Merged
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
Collaborator
CI test resultstest results on build#61767
test results on build#61891
test results on build#61893
test results on build#62179
test results on build#62203
|
Contributor
Author
|
Meh. It worked locally. Will debug. |
- wait for consuming all messages regardless translation state - avoid race conditions when stopping consumer
sometimes maximum throughput is not desired
162ed94 to
3b4aa86
Compare
Contributor
Author
|
/dt |
andrwng
reviewed
Feb 19, 2025
| compacted: bool = False, | ||
| table_override: Optional[str] = None): | ||
| table_override: Optional[str] = None, | ||
| buffer=5000): |
Contributor
There was a problem hiding this comment.
nit: maybe call this max_buffered_msgs or somesuch?
| config=self.avro_stream_config( | ||
| self.TOPIC_NAME, "verifier_schema", 1000000)) | ||
| self.TOPIC_NAME, "verifier_schema", 1000000, | ||
| 1)) |
Contributor
There was a problem hiding this comment.
nit: maybe pull 1 out into some low_interval_ms and explain in a comment why it's necessary?
| verifier = DatalakeVerifier(self.redpanda, | ||
| self.TOPIC_NAME, | ||
| self.dl.spark(), | ||
| buffer=50000) |
Contributor
There was a problem hiding this comment.
nit: maybe pull 50000 out into some high_buffered_msgs and explain in a comment why it's necessary?
| connect.stop_stream("ducky_stream", should_finish=False) | ||
| time.sleep(1) # just it case: let verifier consume remaining messages | ||
| verifier.go_offline() | ||
| verifier.go_offline(600) |
Contributor
There was a problem hiding this comment.
nit: comment explaining why this is necessary?
| include_query_engines=[QueryEngineType.SPARK]) | ||
|
|
||
| def avro_stream_config(self, topic, subject, cnt=3000): | ||
| def avro_stream_config(self, topic, subject, cnt=3000, interval_ms=None): |
Contributor
There was a problem hiding this comment.
Production rate reduced as otherwise RPCN produces too much data while unmounting
Hmm I thought we blocked writes while we unmounted. Are we sure we eventually actually quiesce? The tests I saw were hanging around for 12 minutes without completing -- I don't imagine unmount would take that long, would it?
Contributor
Author
There was a problem hiding this comment.
Yes we do, but it takes some time since we block it.
The 12 minutes you saw was a combination of many problems, swallowed errors in particular -- see 1st commit.
1000 messages per second is not a crazy production rate, but a buffer of 5000 messages will only keep 5 seconds worth, which is less than unmount or translation delays
Production rate reduced as otherwise RPCN produces too much data while unmounting, so it takes unreasonable time to complete. Sleep removed as verifier is more robust. Time limit increased as sometimes unmount takes more time and it takes longer to get offline: - offline mode waits for consuming till migrations blocking offset - consume thread waits for query thread (limited comparison buffer) - query thread may lag because translation lags
3b4aa86 to
323b0b9
Compare
to avoid writing too much data
323b0b9 to
87ce98c
Compare
andrwng
approved these changes
Feb 24, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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.
The test is broken: it swallows errors, timeout ones in particular.
Make it propagate errors. To avoid timeouts:
Backports Required
Release Notes