[
https://issues.apache.org/jira/browse/ZOOKEEPER-288?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12689169#action_12689169
]
Benjamin Reed commented on ZOOKEEPER-288:
-----------------------------------------
looks good. i'd like to see a couple of changes before we commit it:
1) you have genSecurePadding() in two different classes. for maintainability,
you should only have it in one place.
2) in genLedgerKey and genMacKey, why do you allocate a byte array, do the copy
and then do the digest.update(), rather than calling digest.update() twice? i
think it makes is simpler and more readable.
3) MasterKeys should start with a lower case letter. you never remove from
MasterKeys, so you should probably make that a WeakHashMap, so that it can get
GCed.
4) are the masterkeys stored with the ledger fragment? i couldn't find it.
> Cleanup and fixes to BookKeeper
> -------------------------------
>
> Key: ZOOKEEPER-288
> URL: https://issues.apache.org/jira/browse/ZOOKEEPER-288
> Project: Zookeeper
> Issue Type: Improvement
> Components: contrib-bookkeeper
> Affects Versions: 3.0.0
> Reporter: Flavio Paiva Junqueira
> Assignee: Flavio Paiva Junqueira
> Fix For: 3.2.0
>
> Attachments: ZOOKEEPER-BOOKKEEPER-288.patch,
> ZOOKEEPER-BOOKKEEPER-288.patch, ZOOKEEPER-BOOKKEEPER-288.patch
>
>
> We observed one race condition when multiple threads try to write
> concurrently. This patch should fix it. I will also remove some commented
> code.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.