jvrao commented on a change in pull request #1902: (WIP) Metadata checker
validating EnsemblePlacementpolicy
URL: https://github.com/apache/bookkeeper/pull/1902#discussion_r248545565
##########
File path:
bookkeeper-server/src/main/java/org/apache/bookkeeper/replication/Auditor.java
##########
@@ -161,6 +174,12 @@
help = "the number of delayed-bookie-audits cancelled"
)
private final Counter numDelayedBookieAuditsCancelled;
+ @StatsDoc(
+ name =
METADATA_CHECK_ENSEMBLE_NOT_ADHERING_TO_PLACEMENT_POLICY_COUNTER,
+ help = "total number of "
+ + "segments of ledgers failed to adhere to EnsemblePlacementPolicy
found in metadata check"
+ )
Review comment:
What is the advantage we are getting counting each segment rather than one
count per ledger? Just thinking aloud here. If we count by segment, we have no
way of knowing what percentage of the store is wack. @reddycharan do you see
any advantage of counting segments instead of just one increment per ledger?
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
