Joinutility seperatorLogin utility separator Infobright.com
   
 
unexpected segfault
Posted: 03 June 2010 05:59 AM   Ignore ]  
Newbie
Rank
Total Posts:  2
Joined  2010-06-03

Hi,

From time to time, we get a segfault and the Infobright server dies. We don’t really know how to reproduce it and what causes it and the logs doesn’t show any clue about what’s going on.

Is there anything we could do to provide more information so we can find a solution?

We are using Infobright 3.3.1-0 Debian package in a Ubuntu 9.10 64-bit server.

The database is used by our continuous integration system, so it creates and removes many databases quite often, I’m not sure whether this may be the cause.

The message we get from the logs (not very useful…):

messages:Jun 2 12:28:17 london64 kernel: [1785614.589195] mysqld[10292]: segfault at 3a ip 0000000000a00fc8 sp 00007f6760d00450 error 4 in mysqld[400000+8cc000]

Thanks for any help!!

Profile
 
Posted: 03 June 2010 10:13 AM   Ignore ]   [ # 1 ]  
Member
RankRankRank
Total Posts:  218
Joined  2008-08-18

Some suggestions from infrastructure side: enable general query log and slow query log. And try to find if any particular query (based on time of crash and timestamp in query log) that’s making a crash. Set controlmessages to 2 in brighthouse.ini file. Can you copy and paste complete messages from from segfaults log.

Signature 

Mahib

Profile
 
Posted: 03 June 2010 10:31 AM   Ignore ]   [ # 2 ]  
Newbie
Rank
Total Posts:  2
Joined  2010-06-03
Mahib - 03 June 2010 10:13 AM

Some suggestions from infrastructure side: enable general query log and slow query log. And try to find if any particular query (based on time of crash and timestamp in query log) that’s making a crash. Set controlmessages to 2 in brighthouse.ini file. Can you copy and paste complete messages from from segfaults log.


Hi,

We already have all that activated, but we didn’t found anything useful, except that it happens with with ‘LOAD DATA LOCAL INFILE’, but we don’t know what part of the file being loaded is the one causing the break. Is there any log we could activate to see it? or maybe we are not looking at the right place….

Looking at brighthouse.log, we can see:

2010-05-31 19:18:54 Brighthouse engine started.
2010-05-31 19:18:54 2010-05-31 19:18:54 Error - previous session (id=235007344) was not properly committed!
...
2010-05-31 19:18:54 2010-05-31 19:18:54 Error - previous session (id=235007344) was not properly committed!
2010-05-31 19:18:54 2010-05-31 19:18:54 Error - previous session (id=1800254983) was not properly committed!
...
2010-05-31 19:18:54 2010-05-31 19:18:54 Error - previous session (id=1800254983) was not properly committed!
2010-05-31 19:18:54 2010-05-31 19:18:54 Error - previous session (id=780513198) was not properly committed!
...
2010-05-31 19:18:54 2010-05-31 19:18:54 Error - previous session (id=780513198) was not properly committed!
2010-06-01 12:19:06 Brighthouse engine started.
2010-06-01 12:19:06 2010-06-01 12:19:06 Error - previous session (id=1459266079) was not properly committed!
... (around 24 lines like the previous one)
2010-06-01 12:19:06 2010-06-01 12:19:06 Error - previous session (id=1459266079) was not properly committed!
2010-06-01 16:35:13 Brighthouse engine started.


Every “Brighthouse engine started” is at the same time the system log says we got a segfault.

Profile