activemq-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Paul Gale <paul.n.g...@gmail.com>
Subject Re: ActiveMQ 5.8.0: java.io.EOFException: Chunk stream does not exist, page: 19 is marked free
Date Thu, 03 Oct 2013 23:48:55 GMT
Hi Gary,

I'd love to be able to provide a failing unit test for you. However, this
exception occurs infrequently and so far only on one machine. I made a
backup copy of the message store prior to rebuilding it if you think that
examining it might help? It contains only a single 32mb journal file, in
addition to the other files. I'm assuming you have the tooling to be able
to dive in. I wouldn't know where to start. Give me some pointers and I'll
gladly help out.

I suppose I could run some unit test code coverage tools to see if there
are any areas that stand out that lack coverage. What packages should I
zero in on? Is there any support for code coverage analysis in the activemq
build tools? Just wondering.

Thanks,
Paul


On Thu, Oct 3, 2013 at 5:37 PM, Gary Tully <gary.tully@gmail.com> wrote:

> I really thought all of these kahaDB index corruption issues were
> sorted, but if you can reproduce in 5.9 we have a problem that needs
> work. the key to any resolution is a reproducible test case.
>
> In any event, you should be able to recover by rebuilding the index,
> just deleting the db.data file before a restart.
>
> On 3 October 2013 17:04, Paul Gale <paul.n.gale@gmail.com> wrote:
> > Anyone have any some thoughts about this?
> >
> > Thanks,
> > Paul
> >
> > On Tue, Oct 1, 2013 at 12:50 AM, Paul Gale <paul.n.gale@gmail.com>
> wrote:
> >> Hi,
> >>
> >> I'm using ActiveMQ 5.8.0 on RHEL 6.1
> >>
> >> I have noticed the following exception appearing in my broker's log
> >> file. It appears to be related to the once hourly check for expired
> >> messages, as these occur at _exactly_ the same time past the hour, as
> >> given by the expireMessagesPeriod attribute on our topics. I have no
> >> other scheduled checks that go off hourly. Scheduler support is set to
> >> false for the broker (if that matters). The message store is located
> >> on an NFS v3 mount.
> >>
> >> Any thoughts as to what could be causing this or if is this a
> >> well-known issue that's fixed in 5.9.0? I couldn't find anything in
> >> JIRA about this other than AMQ-3906 which was not resolved although
> >> 'Chunk stream does not exist' has cropped up a few times it appears to
> >> have been fixed in earlier releases.
> >>
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.362 | WARN  | Topic
> >>               | Failed to browse Topic: Auction.Event | ActiveMQ
> >> Broker[queue01.qa1] Scheduler
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.362 | java.io.EOFException:
> >> Chunk stream does not exist, page: 19 is marked free
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.362 |   at
> >>
> org.apache.activemq.store.kahadb.disk.page.Transaction$2.readPage(Transaction.java:470)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.362 |   at
> >>
> org.apache.activemq.store.kahadb.disk.page.Transaction$2.<init>(Transaction.java:447)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.362 |   at
> >>
> org.apache.activemq.store.kahadb.disk.page.Transaction.openInputStream(Transaction.java:444)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.363 |   at
> >>
> org.apache.activemq.store.kahadb.disk.page.Transaction.load(Transaction.java:420)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.363 |   at
> >>
> org.apache.activemq.store.kahadb.disk.page.Transaction.load(Transaction.java:377)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.363 |   at
> >>
> org.apache.activemq.store.kahadb.disk.index.BTreeIndex.loadNode(BTreeIndex.java:262)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.363 |   at
> >>
> org.apache.activemq.store.kahadb.disk.index.BTreeIndex.getRoot(BTreeIndex.java:174)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.363 |   at
> >>
> org.apache.activemq.store.kahadb.disk.index.BTreeIndex.iterator(BTreeIndex.java:232)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.363 |   at
> >>
> org.apache.activemq.store.kahadb.MessageDatabase$MessageOrderIndex$MessageOrderIterator.<init>(MessageDatabase.java:2757)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.363 |   at
> >>
> org.apache.activemq.store.kahadb.MessageDatabase$MessageOrderIndex.iterator(MessageDatabase.java:2739)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.363 |   at
> >>
> org.apache.activemq.store.kahadb.KahaDBStore$KahaDBMessageStore$3.execute(KahaDBStore.java:526)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.363 |   at
> >>
> org.apache.activemq.store.kahadb.disk.page.Transaction.execute(Transaction.java:779)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.363 |   at
> >>
> org.apache.activemq.store.kahadb.KahaDBStore$KahaDBMessageStore.recover(KahaDBStore.java:522)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.363 |   at
> >>
> org.apache.activemq.store.ProxyTopicMessageStore.recover(ProxyTopicMessageStore.java:62)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.364 |   at
> >> org.apache.activemq.broker.region.Topic.doBrowse(Topic.java:578)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.364 |   at
> >> org.apache.activemq.broker.region.Topic.access$100(Topic.java:65)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.364 |   at
> >> org.apache.activemq.broker.region.Topic$6.run(Topic.java:703)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.364 |   at
> >>
> org.apache.activemq.thread.SchedulerTimerTask.run(SchedulerTimerTask.java:33)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.364 |   at
> >> java.util.TimerThread.mainLoop(Timer.java:555)
> >> INFO   | jvm 1    | 2013/09/26 05:01:08.364 |   at
> >> java.util.TimerThread.run(Timer.java:505)
> >>
> >> Thanks,
> >> Paul
>
>
>
> --
> http://redhat.com
> http://blog.garytully.com
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message