Feat[STATS]: Partition replication time #909
Open
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
waldgange
force-pushed
the
partition_replication
branch
2 times, most recently
from
f005255 to
7cc82d2
Compare
![bmq-oss-ci[bot]](https://avatars.githubusercontent.com/u/1416818?s=60&v=4){kind=small}
|
@waldgange Can you put in the description what exact metric are you reporting? |
|
Back to Anton for more research |
waldgange
force-pushed
the
partition_replication
branch
2 times, most recently
from
ccbf1cf to
95a3533
Compare
waldgange
force-pushed
the
partition_replication
branch
from
95a3533 to
5d68561
Compare
Signed-off-by: Anton Pryakhin <[email protected]>
waldgange
force-pushed
the
partition_replication
branch
from
5d68561 to
74265e7
Compare
dorjesinpo
reviewed
Oct 16, 2025
There was a problem hiding this comment.
Few thoughts:
- Need to double check if
e_PARTITION_REPLICATIONmetric for the weak consistency justifies the price of extra Receipts traffic. - Rather than introducing extra
StorageMessageAttributesfield, (suppose the answer to the 1. is yes), Replicas can just send Receipt unconditionally. Primary can add PUTs tod_unreceiptedunconditionally (with a new flag, maybe). Primary then can skip callingonReplicatedBatchfor weak consistency messages. - Backward compatibility. Looks like (but need to double-check and test) extra Receipts do not hurt (older versions)
- There is
mqbs::ReplicatedStorage::isStrongConsistencywhich may be useful.
Wrt 1. the consensus is to NOT use the new metric for weak consistency yet. That should reduce the scope of this PR
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Outdated suggestions cannot be applied.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.
This PR introduces a new partition metric
e_PARTITION_REPLICATION_TIME_NS. It represents the time in nanoseconds it takes to store a message record at primary and replicate it to a majority of nodes in the cluster. We report average and maximum time during the report interval only from the primary node. The replication time is measured for both strong and eventual consistency modes. To make it possible we started collecting receipts from replicas regardless of the level of consistency. Also with this PR we stop reporting ACK time from primaries. Now we only do it from "first-hop" nodes