[1.3.6 r4081] Strange message on the server when an agent is restarted

[Farzad FARID]

Hi,

I'm running Zabbix 1.3.6 r4081, the server is on a 64bits Linux FC5, one of the agent is on a 32bits FC5.
After upgrading the agent, while the server was running, I got these messages in the server log:

21545:20070502:173215 Timeout while answering request
21545:20070502:173215 Get value from agent failed. Error: ZBX_TCP_READ() failed [Interrupted system call]
21545:20070502:173215 Host [srv09bc1]: first network error, wait for 15 seconds
21577:20070502:173233 Get value from agent failed. Error: Cannot resolve []
21577:20070502:173233 Host [^_^K<AE><AA>*]: another network error, wait for 15 seconds
21553:20070502:173613 Get value from agent failed. Error: Cannot connect to [srv09bc1.mydomain.com:10050] [Connection refused]
21553:20070502:173613 Host [srv09bc1]: first network error, wait for 15 seconds

What worries me are the lines:

• 21577:20070502:173233 Get value from agent failed. Error: Cannot resolve []
• 21577:20070502:173233 Host [^_^K<AE><AA>*]: another network error, wait for 15 seconds

This looks like a memory corruption or a bad pointer. As I said previously the zabbix server is running on a 64 bits Linux.

Regards.

[Farzad FARID]

Another corrupted message on linux 64 bits

Hi,

Last night my server running Zabbix Server 1.3.6 r4081, on a 64 bits Fedora Core 5, show again a strange log message that seems to results from corrupted memory or a stray pointer. This time it happened in a different string:

 21545:20070502:200315 Timeout while answering request
 21545:20070502:200315 Get value from agent failed. Error: ZBX_TCP_READ() failed [Interrupted system call]
 21545:20070502:200315 Host [srv09bc1]: first network error, wait for 15 seconds
 21577:20070502:200332 Delay period format is wrong [9^Wc]
 21574:20070502:200938 Executing housekeeper
 21574:20070502:200939 Deleted 1683 records from history and trends

Regards.

[Farzad FARID]

Hi,

Another corrupted message, still on the server, after I restarted an agent that didn't work anymore:

 21577:20070503:101709 Get value from agent failed. Error: ZBX_TCP_READ() failed [Interrupted system call]
 21577:20070503:101709 Host [srv05bc2] will be checked after 60 seconds
 21567:20070503:101729 Query::select i.key_,i.delay,i.lastlogsize from items i,hosts h where i.hostid=h.hostid and h.status=0 and i.status
=0 and i.type=7 and h.host='srv05bc2' and h.hostid>=100000000000000*0 and h.hostid<=(100000000000000*0+99999999999999)
 21567:20070503:101729 Query failed:MySQL server has gone away [2006]
 21577:20070503:101809 Delay period format is wrong [^Wc]
 21577:20070503:101809 Enabling host [^Vc]

By the way, you can notice the worrying message Query failed:MySQL server has gone away, whereas the mysql server actually works correctly. I don't know why restarted a non-functionnal agent triggered this bug.

And the agent but is one I got with zabbix 1.3.5, and that is still not corrected in zabbix 1.3.6 r4081 : http://www.zabbix.com/forum/showthread.php?t=5829

After this the freshly restarted agent and the server seem to work correctly again.

Regards.

[Alexei]

Thanks for reporting this. It looks like some data corruption insider ZABBIX server. To be fixed.

[Alexei]

Indeed, these messages were related to usage of freed structure. Fixed!