[
https://issues.apache.org/jira/browse/HBASE-9098?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13727228#comment-13727228
]
Ted Yu commented on HBASE-9098:
-------------------------------
{code}
- // when a region is in recovering state, it can only accept writes not reads
- private volatile boolean recovering = false;
{code}
Nice simplification above.
{code}
- public void prepareLogReplay(Set<ServerName> serverNames) throws IOException
{
+ public void prepareLogReplay(ServerName serverName, List<HRegionInfo>
regions) throws IOException {
{code}
There is conversion later in the patch:
{code}
+ for(HRegionInfo tmpRegionInfo : regions) {
+ regionSet.add(tmpRegionInfo);
{code}
Why not passing Set<HRegionInfo> as parameter ?
{code}
+ // verify current region is indeed in recovering state
+ try {
+ if (SplitLogManager.isRegionMarkedRecoveringInZK(watcher,
loc.getRegionInfo()
+ .getEncodedName()) == false) {
{code}
The comment seems to be different from the check above.
> During recovery use ZK as the source of truth for region state
> ---------------------------------------------------------------
>
> Key: HBASE-9098
> URL: https://issues.apache.org/jira/browse/HBASE-9098
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Affects Versions: 0.95.0
> Reporter: Devaraj Das
> Assignee: Jeffrey Zhong
> Priority: Blocker
> Fix For: 0.95.2
>
> Attachments: hbase-9098.patch
>
>
> In HLogSplitter:locateRegionAndRefreshLastFlushedSequenceId(HConnection,
> byte[], byte[], String), we talk to the replayee regionserver to figure out
> whether a region is in recovery or not. We should look at ZK only for this
> piece of information (since that is the source of truth for recovery
> otherwise).
--
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
