[
https://issues.apache.org/jira/browse/CASSANDRA-5612?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13741526#comment-13741526
]
Aleksey Yeschenko commented on CASSANDRA-5612:
----------------------------------------------
So, what's happening:
1. node2 comes up, slightly earlier than node3
2. node2 calls Auth.setupAuthKeyspace(); but node3 is not alive yet, so this
change is *not* pushed to node3
3. node3 comes up and gets noticed by node2
4. node2 calls Auth.setupUsersTable(); the new CF is pushed to all the nodes,
including node3. However, node3 does *not* have the keyspace itself, so NPE is
thrown at
https://github.com/apache/cassandra/blob/cassandra-1.2/src/java/org/apache/cassandra/db/DefsTable.java#L514
- because ksm is null
> NPE when upgrading a mixed version 1.1/1.2 cluster fully to 1.2
> ---------------------------------------------------------------
>
> Key: CASSANDRA-5612
> URL: https://issues.apache.org/jira/browse/CASSANDRA-5612
> Project: Cassandra
> Issue Type: Bug
> Affects Versions: 1.2.6
> Reporter: Ryan McGuire
> Assignee: Aleksey Yeschenko
> Attachments: logs.tar.gz, upgrade_through_versions_test.py
>
>
> See the attached upgrade_through_versions_test.py upgrade_test_mixed().
> Conceptually this method does the following:
> * Instantiates a 3 node 1.1.9 cluster
> * Writes some data
> * Shuts down node 1 and upgrades it to 1.2 (HEAD)
> * Brings the node1 back up, making the cluster a mixed version 1.1/1.2
> * Brings down node2 and node3 and does the same upgrade making it all the
> same version.
> * At this point, I would run upgradesstables on each of the nodes, but there
> is already an error on node3 directly after it's upgrade:
> {code}
> INFO [FlushWriter:1] 2013-06-03 22:49:46,543 Memtable.java (line 461) Writing
> Memtable-peers@1023263314(237/237 serialized/live bytes, 14 op
> s)
> INFO [FlushWriter:1] 2013-06-03 22:49:46,556 Memtable.java (line 495)
> Completed flushing /tmp/dtest-YqMtHN/test/node3/data/system/peers/syst
> em-peers-ic-2-Data.db (291 bytes) for commitlog position
> ReplayPosition(segmentId=1370314185862, position=58616)
> INFO [GossipStage:1] 2013-06-03 22:49:46,568 StorageService.java (line 1330)
> Node /127.0.0.2 state jump to normal
> ERROR [MigrationStage:1] 2013-06-03 22:49:46,655 CassandraDaemon.java (line
> 192) Exception in thread Thread[MigrationStage:1,5,main]
> java.lang.NullPointerException
> at
> org.apache.cassandra.db.DefsTable.addColumnFamily(DefsTable.java:511)
> at
> org.apache.cassandra.db.DefsTable.mergeColumnFamilies(DefsTable.java:445)
> at org.apache.cassandra.db.DefsTable.mergeSchema(DefsTable.java:355)
> at
> org.apache.cassandra.db.DefinitionsUpdateVerbHandler$1.runMayThrow(DefinitionsUpdateVerbHandler.java:55)
> at
> org.apache.cassandra.utils.WrappedRunnable.run(WrappedRunnable.java:28)
> at
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:471)
> at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:334)
> at java.util.concurrent.FutureTask.run(FutureTask.java:166)
> at
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
> at java.lang.Thread.run(Thread.java:722)
> {code}
> This error is repeatable, but inconsistent. Interestingly, it is always node3
> with the error.
--
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
