Zabbix server crashes after upgrading to 7.2

[clevelas]

We upgraded from 6.4 to 7.2 a couple of weeks ago. Ever since, we have been having crashes. Things like "double free or corruption (fasttop)". At first it was once a day, now it's every few days (we tweaked some poller numbers that seemed to help some). I'm not looking for an answer right now, but asking what I need to do to gather more information to fix the issue.

Server version: 7.2.5

Any help is appreciated.

- Steve

[clevelas]

Here are some examples:

On 7.2.4, I think, we had hard crashes that stopped zabbix server from functioning. Processes were still running and were difficult to kill. The logs looked like this:

double free or corruption (fasttop)
2725522:20250326:230444.539 Got signal [signal:11(SIGSEGV),reason:1,refaddr:(nil)]. Crashing ...

I don't think we've had it hard crash in 7.2.5 yet, but it has had processes die and the server restart itself. Here is the example from yesterday:

​double free or corruption (fasttop)
2684061:20250414:141616.110 One child process died (PID:2684105,exitcode/signal:6). Exiting ...
2684062:20250414:141616.110 HA manager has been paused
zabbix_server [2684061]: Error waiting for process with PID 2684105: [10] No child processes
2684062:20250414:141616.208 HA manager has been stopped
2684061:20250414:141616.256 syncing history data...
2684061:20250414:141616.259 syncing history data... 46.153846%
2684061:20250414:141616.259 syncing history data done
2684061:20250414:141616.358 syncing trend data in progress...
2684061:20250414:141618.918 syncing trend data... 100.000000%
2684061:20250414:141618.918 syncing trend data done
2684061:20250414:141619.042 Zabbix Server stopped. Zabbix 7.2.5 (revision d2eb5068f4b).
2709777:20250414:141629.091 Starting Zabbix Server. Zabbix 7.2.5 (revision d2eb5068f4b).

It's crashed like this at least three times since we upgraded to 7.2.5. I'm still looking for some suggestions on how to collect more data on these crashes.

Thanks,
- Steve

[Isopote42]

Hello,

Same problem here, since 7.2

520285:20250420:093748.661 thread started
double free or corruption (fasttop)
malloc_consolidate(): unaligned fastbin chunk detected
519975:20250420:140736.693 Got signal [signal:11(SIGSEGV),reason:1,refaddr:(nil)]. Crashing ...
519975:20250420:140736.693 ====== Fatal information: ======
519975:20250420:140736.693 Program counter: 0x7f86a0a994b9
519975:20250420:140736.693 === Registers: ===
519975:20250420:140736.693 r8      =     7f862000b6a0 =      140214039262880 =      140214039262880
519975:20250420:140736.693 r9      =                0 =                    0 =                    0
519975:20250420:140736.693 r10     =                8 =                    8 =                    8
519975:20250420:140736.693 r11     =     7f8620000020 =      140214039216160 =      140214039216160

I try to set "DebugLevel" to more verbose but lot of log is difficult for the storage...