[
https://issues.apache.org/jira/browse/BOOKKEEPER-293?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13532240#comment-13532240
]
Ivan Kelly commented on BOOKKEEPER-293:
---------------------------------------
[~umamahesh], the amount of traffic to bookies for these checks is minimal. For
each ledger we read 2 entries on each bookie in each ensemble. In the normal
case this will amount to reading 6 entries per ledger, though one issue with it
could be that it would cause a evictions in the cache, but in terms of network
bandwidth, it's not an issue.
What is an issue, and something that we'll have to look at in 4.3.0 is that if
there are a lot of ledgers, the periodic checker will end up accessing
zookeeper a lot, but this is a problem for the auditor in general. We should
see if we can find some way to a) reduce the load on zk and b) spread the reads
on zookeeper around the cluster.
> Periodic checking of ledger replication status
> ----------------------------------------------
>
> Key: BOOKKEEPER-293
> URL: https://issues.apache.org/jira/browse/BOOKKEEPER-293
> Project: Bookkeeper
> Issue Type: Sub-task
> Components: bookkeeper-auto-recovery
> Reporter: Ivan Kelly
> Assignee: Ivan Kelly
> Fix For: 4.2.0
>
> Attachments:
> 0001-BOOKKEEPER-293-Periodic-checking-of-ledger-replicati.patch,
> 0002-BOOKKEEPER-293-Periodic-checking-of-ledger-replicati.patch
>
>
> We should periodically check all ledgers to make sure that they are fully
> replicated.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
