-
Problem report
-
Resolution: Unresolved
-
Trivial
-
None
-
None
-
None
I have noticed continuous memory usage growth in proxy pollers when using 'Veeam Backup Enterprise Manager by HTTP' template. After removing templates memory usage is not decreasing and you need to restart proxy service. Seems like it is only happening when veeam manager api query is not working or causing timeout.
I have similar test environment with same zabbix version and there everything seems to work OK.
Tested with proxy 6.0.30, ubuntu 20.04 / 8GB RAM
Steps to reproduce:
- Add hosts with 'Veeam Backup Enterprise Manager by HTTP' template.
- item: Veeam Manager: Get errors
Error: cannot get URL: Timeout was reached.
Error: script execution timeout occurred.
Result:
Before adding hosts and veeam template
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
zabbix 81025 0.0 0.2 1731528 22236 ? S Jul05 0:42 /usr/sbin/zabbix_proxy: poller #28 [got 0 values in 0.000020 sec, idle 1 sec]
zabbix 81035 0.0 0.2 1731532 22224 ? S Jul05 0:41 /usr/sbin/zabbix_proxy: poller #38 [got 0 values in 0.000008 sec, idle 1 sec]
zabbix 80994 0.0 0.2 1731520 22212 ? S Jul05 0:40 /usr/sbin/zabbix_proxy: poller #3 [got 0 values in 0.000021 sec, idle 1 sec]
zabbix 81026 0.0 0.2 1731676 22208 ? S Jul05 0:40 /usr/sbin/zabbix_proxy: poller #29 [got 0 values in 0.000010 sec, idle 1 sec]
zabbix 81028 0.0 0.2 1731676 22208 ? S Jul05 0:43 /usr/sbin/zabbix_proxy: poller #31 [got 0 values in 0.000010 sec, idle 1 sec]
*SNIP
After sometime (in this case 20 hours):
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
zabbix 80999 0.0 1.2 1815884 103664 ? S Jul05 1:04 /usr/sbin/zabbix_proxy: poller #7 [got 0 values in 0.000006 sec, idle 5 sec]
zabbix 81056 0.0 1.1 1802472 90156 ? S Jul05 1:11 /usr/sbin/zabbix_proxy: poller #59 [got 0 values in 0.000024 sec, idle 5 sec]
zabbix 81005 0.0 1.1 1801764 89816 ? S Jul05 1:12 /usr/sbin/zabbix_proxy: poller #13 [got 0 values in 0.000033 sec, idle 5 sec]
zabbix 81035 0.0 1.1 1801728 89732 ? S Jul05 1:06 /usr/sbin/zabbix_proxy: poller #38 [got 0 values in 0.000020 sec, idle 5 sec]
*SNIP