Segfault Zabbix-agentd 1.1.4-2 Debian

[zabbix-emaxx]

Every time I restart zabbix_agentd I get this on the console and in syslog:

Jan  5 10:33:13 xxx kernel: zabbix_agentd[2694]: segfault at 0000000000000000 rip 00002b352558a8d8 rsp 00007fff85a68de0 error 4
Jan  5 10:48:09 xxx kernel: zabbix_agentd[2746]: segfault at 0000000000000000 rip 00002b28db9988d8 rsp 00007fffcf65aa30 error 4

System and package info:

Linux xxx 2.6.17.13-vs2.0.2.1.20070102 #1 SMP Tue Jan 2 16:08:52 UTC 2007 x86_64 GNU/Linux

Debian GNU/Linux 4.0 Etch/Testing

Package: zabbix-agent
Priority: optional
Section: net
Installed-Size: 376
Maintainer: Zabbix Maintainers <[email protected]>
Architecture: amd64
Source: zabbix
Version: 1:1.1.4-2
Depends: libc6 (>= 2.3.5-1), libldap2 (>= 2.1.17-1), debconf (>= 0.5) | debconf-2.0, adduser, logrotate
Filename: pool/main/z/zabbix/zabbix-agent_1.1.4-2_amd64.deb
Size: 121576
MD5sum: b0592076cfbfd6bc90e27370b8dc5cc7
SHA1: 4e16e25ae57b4ff0311dd7241b20d877d6440b7f
SHA256: fa389e4d7010d7a27f327ad81444094a5a5ffea4baadcf0a61e693be60d51e27

Only the zabbix-agent seems to have this segfault, have not seen other software do this, so the system hardware should be ok(?).

Zabbix seems to work fine with my Zabbix-1.1 Server (not on the same machine)

Anyone got a clue what is causing this segfault?

[abi]

hi,

i cant reproduce those segfaults. Can you provide some logfiles with
debugging set to 4?

[Alexei]

I would be also very interested to hear if this is a Debian specific issue or some general problem.

[abi]

I would be also very interested to hear if this is a Debian specific issue or some general problem.

might also be a issue in the kernel, though (these types of segfaults
are related to the kernel, afaik).

Could you try to update your (outdated) 2.6.17 to the lastest 2.6.18-3
Packages (there are packages with vserver support if you need that)
and see if these segfaults still happen?

As said, i cant reproduce it on my amd64(unstable) or x86(testing) box.
I've also tried with old debian 2.6.17-vserver packages and couldnt
reproduce it either (x86).

[zabbix-emaxx]

[QOUTE=Alexei]i cant reproduce those segfaults. Can you provide some logfiles with debugging set to 4?[/QUOTE]

might also be a issue in the kernel, though (these types of segfaults
are related to the kernel, afaik).

Could you try to update your (outdated) 2.6.17 to the lastest 2.6.18-3
Packages (there are packages with vserver support if you need that)
and see if these segfaults still happen?

As said, i cant reproduce it on my amd64(unstable) or x86(testing) box.
I've also tried with old debian 2.6.17-vserver packages and couldnt
reproduce it either (x86).

I will try that, but - to be sure - I am running Memtest86+ v1.65 first and it already discoverd two memory(?) problems (zie attached image).

(still didn't have problems with other software ...strange)

[alj]

I would be also very interested to hear if this is a Debian specific issue or some general problem.

I run zabbix-agent on debian-etch on about 200 machines now and didn't see segfault yet

It is dying however during log rotation being blocked on IO.

So i had to disable logrotate for zabbix completely (remove /etc/logrotate.d/zabbix-agent file) and agent is working fine now.

[zabbix-emaxx]

might also be a issue in the kernel, though (these types of segfaults
are related to the kernel, afaik).

Could you try to update your (outdated) 2.6.17 to the lastest 2.6.18-3
Packages (there are packages with vserver support if you need that)
and see if these segfaults still happen?

As said, i cant reproduce it on my amd64(unstable) or x86(testing) box.
I've also tried with old debian 2.6.17-vserver packages and couldnt
reproduce it either (x86).

Hmmzzz .. I got an other machine with the exact same setup, but with the Debian Etch/Testing standard kernel (without vserver support) giving the same segfault:

Jan  6 00:46:33 yyy kernel: zabbix_agentd[6532]: segfault at 0000000000000000 rip 00002b8ced7ee8d8 rsp 00007fffbd804b80 error 4

System info:

Linux yyy 2.6.18-3-amd64 #1 SMP Mon Dec 4 17:04:37 CET 2006 x86_64 GNU/Linux

processor       : 0
vendor_id       : GenuineIntel
cpu family      : 15
model           : 6
model name      :               Intel(R) Pentium(R) D CPU 2.80GHz
stepping        : 4
cpu MHz         : 2800.007
cache size      : 2048 KB
physical id     : 0
siblings        : 2
core id         : 0
cpu cores       : 2
fpu             : yes
fpu_exception   : yes
cpuid level     : 6
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall nx lm constant_tsc pni monitor ds_cpl est cid cx16 xtpr lahf_lm
bogomips        : 5606.08
clflush size    : 64
cache_alignment : 128
address sizes   : 36 bits physical, 48 bits virtual
power management:

processor       : 1
vendor_id       : GenuineIntel
cpu family      : 15
model           : 6
model name      :               Intel(R) Pentium(R) D CPU 2.80GHz
stepping        : 4
cpu MHz         : 2800.007
cache size      : 2048 KB
physical id     : 0
siblings        : 2
core id         : 1
cpu cores       : 2
fpu             : yes
fpu_exception   : yes
cpuid level     : 6
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall nx lm constant_tsc pni monitor ds_cpl est cid cx16 xtpr lahf_lm
bogomips        : 5599.92
clflush size    : 64
cache_alignment : 128
address sizes   : 36 bits physical, 48 bits virtual
power management:

What info can I give you more to reproduce the problem?

PS: I did run MemTest86+ on this system for about 2 hours, I got no errors.

[zabbix-emaxx]

What info can I give you more to reproduce the problem?

If I can do some more testing, please let me know.

[zabbix-emaxx]

If I can do some more testing, please let me know.

I didn't see any segfaults anymore ... strange.

Guess the problem dissapeared somehow (did not change any kernel or hardware).

[yankeedoodle]

Confirmation of Issue

I have a cluster a dozen dual-opterons that I'm using for a test environment running the latest build of Debian etch, and I'm seeing the same issue duplicated across all machines.

I tried two different debian zabbix-agent versions and the error is still replicated.

zabbix-agent_1.1.4-7_amd64.deb, Segfault:

zabbix_agentd[14679]: segfault at 0000000000000000 rip 00002b4ee491b8d8 rsp 00007fffc66d5a60 error 4

zabbix-agent_1%3a1.1.4-2_amd64.deb, Segfault:

zabbix_agentd[3042]: segfault at 0000000000000000 rip 00002ba2af3f98d8 rsp 00007ffffbbf9a70 error 4

lab1:~# uname -a
Linux lab1 2.6.18-3-amd64 #1 SMP Mon Dec 4 17:04:37 CET 2006 x86_64 GNU/Linux

I'm happy to run simple tests to help isolate the problem, if instructions are provided.

[Alexei]

Does this problem occur with Debian agent only? Can you reproduce it with agent binary built from ZABBIX sources?