[
https://issues.apache.org/jira/browse/HDFS-7934?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14362978#comment-14362978
]
Vinayakumar B commented on HDFS-7934:
-------------------------------------
I think the problematic area is below code block in Fsimage#loadFsImage(..)
{code} // For rollback in rolling upgrade, we need to set the
toAtLeastTxId to
// the txid right before the upgrade marker.
long toAtLeastTxId = editLog.isOpenForWrite() ? inspector
.getMaxSeenTxId() : 0;
if (rollingRollback) {
// note that the first image in imageFiles is the special checkpoint
// for the rolling upgrade
toAtLeastTxId = imageFiles.get(0).getCheckpointTxId() + 2;
}{code}
In Case of rollingRollback, there is nothing read from edits streams. So
setting {{toAtLeastTxId = imageFiles.get(0).getCheckpointTxId() + 2;}} is not
required. removing this line will solve the problem IMO.
Any thoughts?
> During Rolling upgrade rollback ,standby namenode startup fails.
> ----------------------------------------------------------------
>
> Key: HDFS-7934
> URL: https://issues.apache.org/jira/browse/HDFS-7934
> Project: Hadoop HDFS
> Issue Type: Bug
> Reporter: J.Andreina
> Assignee: J.Andreina
> Priority: Critical
>
> During Rolling upgrade rollback , standby namenode startup fails , while
> loading edits and when there is no local copy of edits created after upgrade
> ( which is already been removed by Active Namenode from journal manager and
> from Active's local).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
