activemq-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Lionel Cons (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (APLO-284) Invalid log position warnings
Date Fri, 01 Feb 2013 07:33:13 GMT

    [ https://issues.apache.org/jira/browse/APLO-284?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13568540#comment-13568540
] 

Lionel Cons commented on APLO-284:
----------------------------------

Still no luck in reproducing it but it happens again this morning on a broker freshly upgraded
to apollo-99-trunk-20130131.193357-177.

We see many messages like:

2013-02-01 08:06:33,656 | WARN  | Queue 'xxx' detected store dropped message at seq: 1511606
| 
2013-02-01 08:06:33,658 | WARN  | Queue 'xxx' detected store dropped message at seq: 1511607
| 
2013-02-01 08:06:33,659 | WARN  | Queue 'xxx' detected store dropped message at seq: 1511608
| 

But also store compaction messages:

2013-02-01 08:09:14,047 | INFO  | Compacting the leveldb index at: /var/lib/apollo/data/dirty.index
| 
2013-02-01 08:09:14,204 | INFO  | Compaction completed | 
2013-02-01 08:09:24,049 | INFO  | Compacting the leveldb index at: /var/lib/apollo/data/dirty.index
| 
2013-02-01 08:09:24,206 | INFO  | Compaction completed | 
2013-02-01 08:09:34,050 | INFO  | Compacting the leveldb index at: /var/lib/apollo/data/dirty.index
| 
2013-02-01 08:09:34,242 | INFO  | Compaction completed | 

So the weird warning plus the too frequent store compaction seem linked. FWIW, this broker
only serves 19 clients sending each very few messages per second. No big scale, no stress
test.

The producers send to a topic and there are 2 dsubs on it. There is 1 consumers per dsub:
one is getting its messages, the other one is stuck, active but getting no messages (so they
are accumulating in the dsub). The web console shows:

enqueued: 29,316 messages [so there are messages]
consumers: 1 [so there is a consumer]
dispatched: 0 messages [but it gets nothing]
next message seq: 1,845,821
acquired: 0 messages (0 bytes)
acks: 0 messages
naks: 0 messages
waiting on: dispatch

I don't know what "waiting on: dispatch" means...
                
> Invalid log position warnings
> -----------------------------
>
>                 Key: APLO-284
>                 URL: https://issues.apache.org/jira/browse/APLO-284
>             Project: ActiveMQ Apollo
>          Issue Type: Bug
>         Environment: apollo-99-trunk-20121221.031630-148
>            Reporter: Lionel Cons
>
> I don't know if this is related to APLO-176 but on a broker running apollo-99-trunk-20121221.031630-148
we see many (>1k) warnings like:
> 2013-01-07 10:02:55,941 | WARN  | Invalid log position: 122526748304 | 
> 2013-01-07 10:02:55,941 | WARN  | Invalid log position: 122526748923 | 
> 2013-01-07 10:02:55,942 | WARN  | Invalid log position: 122526750644 | 
> 2013-01-07 10:02:55,942 | WARN  | Invalid log position: 122526751727 | 
> 2013-01-07 10:02:55,942 | WARN  | Invalid log position: 122526752423 | 
> Does this indicate some kind of data store corruption?

--
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

Mime
View raw message