[
https://issues.apache.org/jira/browse/HBASE-15682?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15366974#comment-15366974
]
Ted Yu commented on HBASE-15682:
--------------------------------
{code}
+ StringUtils.join(incrBackupFileList, ",") +" to "
{code}
public static String join(CharSequence separator, Iterable<?> strings) {
Separator goes first.
{code}
+ List<String> converted = convertFilesFromWALtoOldWAL(missingFiles);
{code}
The above is called twice: once inside updateIncrBackupFileList() and once
after calling updateIncrBackupFileList().
See if the return value can be reused.
{code}
+ for(String path: missingFiles){
+ if(path.indexOf(Path.SEPARATOR + HConstants.HREGION_LOGDIR_NAME) < 0) {
{code}
nit: insert space between for and (.
{code}
+ while(!success){
{code}
Should there be limit on the number of retries ?
> HBase Backup Phase 3: Possible data loss during incremental WAL files copy
> --------------------------------------------------------------------------
>
> Key: HBASE-15682
> URL: https://issues.apache.org/jira/browse/HBASE-15682
> Project: HBase
> Issue Type: Bug
> Affects Versions: 2.0.0
> Reporter: Vladimir Rodionov
> Assignee: Vladimir Rodionov
> Labels: backup
> Fix For: 2.0.0
>
> Attachments: HBASE-15682-v2.patch
>
>
> We collect list of files in WAL and oldWALs directory and launch DistCp job.
> Some files can be moved from WALs to oldWALs directory by RS during job's
> run, what can result in potential data loss.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
