Zabbix server crashes

[hookproto]

Hello,
According to Colizab, I changed innodb_buffer_pool_size to 8G (I have 32G RAM on this server), without any problem when restarting.

Hookproto, what's the point ? It's over my knowledge...

Thanks
N

We are both experiencing a SIGTERM that fails to terminate the Zabbix process, indicating that the process is stuck in an uninterruptible state.

Both of us are running Debian 12.

• You encountered this issue on Zabbix Server 7.0.9, while I observed it on Zabbix Server 7.2.3.
• Additionally, I noticed the same behavior on Zabbix Proxy 6.4.21.

Since these are the latest versions of their respective branches, it suggests that a recent change is responsible. Based on the changelog, the only fix applied across all three versions is ZBX-25865.

[BenjaminG]

We are both experiencing a SIGTERM that fails to terminate the Zabbix process, indicating that the process is stuck in an uninterruptible state.

Both of us are running Debian 12.

• You encountered this issue on Zabbix Server 7.0.9, while I observed it on Zabbix Server 7.2.3.
• Additionally, I noticed the same behavior on Zabbix Proxy 6.4.21.

Since these are the latest versions of their respective branches, it suggests that a recent change is responsible. Based on the changelog, the only fix applied across all three versions is ZBX-25865.

FYI: I do not use vmware monitoring. So at least for my case, that fix should not be part of the problem

[cwachs]

I am seeing this as well. I am running Ubuntu 24.04 with mysql and Zabbix 7.2.3. Saw crashes on 7.2.2 as well. I'll try the buffer pool size modification but that feel slike we are just kicking the can further down the road...

[JuergenM]

Same here. Running 3 instances (2 x 7.0.9, 1 x 7.2.3, all Ubuntu 24.04 with MySql). Reboot not possible, because Zabbix server never stops. Only hard reset on all installations (including proxies)! I don't think it has something to do with mysql buffer as there is one realy tiny installation (30 vps) and all three show the same symptom.

[rafaaall]

We have the same problem.
Debian 12.9, Zabbix 7.0.9, MariaDB (upgrading to 11.6.2+maria~deb12 didn’t help).
We are capturing packets to the database (localhost) to investigate.
It seems that sometimes, when Zabbix executes an INSERT query (insert into history), MySQL does not send back a response code (0x00 - OK), causing the connection to block other queries.
When the problem is relatively new, killing the INSERT process in MySQL allows other queries to resume, and Zabbix starts working again.

[svc_zbx_6754]

We have also noticed this problem. According to our analysis, it has been happening to us since the upgrade to 7.0.8 - same issues on 7.0.9. Most of the time, one or two proxies fail, but once even the server went down.

[Bellule16]

I also have the same problem.
it has been happening to us since the upgrade to 7.0.6 - same issues on 7.0.9.

[Timej26]

Same for us since 7.0.7. Server crashes, proxies are freezing and need a hard reset at random times. No errors in logs, because the assets freeze and logging stops.

[nicolasfo]

Hello all,
3 days without crash, I just increased max_connections to 200 in /etc/mysql/mariadb.conf.d/50-server.cnf
Don't know if it will stay in long term...
I monitor 320 hosts
N

[eticad]

Hi,
same problem here with new install and 80 hosts.
Tried the max_connections fix.
Thanks

[eticad]

Update: problem fixed with /etc/mysql/mariadb.conf.d/50-server.cnf innodb_buffer_pool_size = 6G

[robin_bolli]

Hi guys, i have the same problem. I changed from Debian to Ubuntu and the problem was solved. I don't know why, but it worked out for me. The max_connections fix didn't worked out for me.

[robin_bolli]

Short addendum to my post: It only works when you use Ubuntu with the graphical interface. If you use only the Ubuntu Server without GUI, I get the same problem as before with Debian. I monitor approximately 350 hosts.

[robin_bolli]

Update: solved with CacheSize=1G instead of 32M

[al3ma]

Hi, I had the same problem during the last few weeks after upgrading to 7.2.3, and now it seems stable again.
The last thing I did was upgrade the templates (from 6.4.0 to 7.2).
I'm not sure if it's linked, but I haven't seen a deadlock since.

[BenjaminG]

FYI
I have gradually increased my buffer and haven't ad a crash since I have
innodb_buffer_pool_size = 8G
So if you have this issue, maybe give it some more cache, depending on your environment size