[
https://issues.apache.org/jira/browse/BOOKKEEPER-299?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13398281#comment-13398281
]
Ivan Kelly commented on BOOKKEEPER-299:
---------------------------------------
I would prefer to keep the actual rereplication logic in the client package as
it does reach into internals which we would have to make public otherwise.
Moving it out of BookKeeperAdmin isn't a bad idea though, but the API should
say on BookKeeperAdmin and call whatever the new class is. Pretty much take
LedgerFragmentReplicator, and move it into client package.
Another thing which would be nice to change here, is that currently the
recovery code will try to recover all entries in parallel. I don't like this. I
think it would be better to do it in series, so that we don't overload the
server with a lot of read requests at once.
> Provide LedgerFragmentReplicator which should replicate the fragments found
> from LedgerChecker
> ----------------------------------------------------------------------------------------------
>
> Key: BOOKKEEPER-299
> URL: https://issues.apache.org/jira/browse/BOOKKEEPER-299
> Project: Bookkeeper
> Issue Type: Sub-task
> Components: bookkeeper-client, bookkeeper-server
> Affects Versions: 4.2.0
> Reporter: Uma Maheswara Rao G
> Assignee: Uma Maheswara Rao G
> Attachments: BOOKKEEPER-299.patch
>
>
> Replication worker requires LedgerFragmentReplicator for replicating the
> actula fragments found from Ledger checker.
> Most of the fragment replication code available in BookKeeperAdmin. We can
> refactor it to LedgerFragmentReplicator and use it.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
