-
Type:
Incident report
-
Resolution: Unresolved
-
Priority:
Trivial
-
None
-
Affects Version/s: None
-
Component/s: None
-
None
I noticed a strange case. I have two Zabbix proxy hosts running in Docker.
Proxy version: /operator-zabbix-proxy:7.2.3
Zabbix version: 7.2.4
10 SNMP pollers are running on each proxy
ps auxw|grep snmp\ poller 1997 2813881 0.0 0.0 709156 3896 ? Sl 10:25 0:00 /usr/sbin/zabbix_proxy: snmp poller #1 [got 0 values, queued 0 in 5 sec, awaiting 0] 1997 2813882 29.7 0.2 752624 31824 ? Rl 10:25 1:58 /usr/sbin/zabbix_proxy: snmp poller #2 [got 93 values, queued 101 in 5 sec, awaiting 53] 1997 2813883 0.0 0.0 709156 3896 ? Sl 10:25 0:00 /usr/sbin/zabbix_proxy: snmp poller #3 [got 0 values, queued 0 in 5 sec, awaiting 0] 1997 2813885 0.0 0.0 709156 3896 ? Sl 10:25 0:00 /usr/sbin/zabbix_proxy: snmp poller #4 [got 0 values, queued 0 in 5 sec, awaiting 0] 1997 2813887 0.1 0.1 724792 23484 ? Sl 10:25 0:00 /usr/sbin/zabbix_proxy: snmp poller #5 [got 0 values, queued 0 in 5 sec, awaiting 0] 1997 2813888 0.0 0.0 709156 3896 ? Sl 10:25 0:00 /usr/sbin/zabbix_proxy: snmp poller #6 [got 0 values, queued 0 in 5 sec, awaiting 0] 1997 2813889 1.3 0.2 724928 26216 ? Sl 10:25 0:05 /usr/sbin/zabbix_proxy: snmp poller #7 [got 0 values, queued 0 in 5 sec, awaiting 0] 1997 2813890 0.1 0.1 724784 23464 ? Sl 10:25 0:00 /usr/sbin/zabbix_proxy: snmp poller #8 [got 0 values, queued 0 in 5 sec, awaiting 0] 1997 2813891 0.0 0.0 709156 3896 ? Sl 10:25 0:00 /usr/sbin/zabbix_proxy: snmp poller #9 [got 0 values, queued 0 in 5 sec, awaiting 0] 1997 2813893 0.1 0.1 724708 23464 ? Sl 10:25 0:00 /usr/sbin/zabbix_proxy: snmp poller #10 [got 0 values, queued 0 in 5 sec, awaiting 0]
After restarting the containers, more pollers work for a while, for example 5 of them. In this case, 5-6 of them process data for 22 seconds.
They process data from the queue:
$ while : ; do sleep 1; ps auxw|grep -i snmp|awk '$16 > 0 && $14 != "#7" {print $14}'; done #10 #2 #3 #10 #1 #2 #3 #6 #10 #1 #2 #3 #6 #10 #1 #2 #3 #6 #10 #1 #3 #6 #6 #6 #6 #6
...
After this time, the data is processed by only one poller (here #6).
Its queue increases again:
$ while : ; do sleep 1; ps auxw|grep -i snmp|grep \#6; done 1997 62317 25.7 0.2 1250020 30588 ? Sl 10:23 3:21 /usr/sbin/zabbix_proxy: snmp poller #6 [got 83 values, queued 82 in 5 sec, awaiting 49] 1997 62317 25.6 0.2 1250020 30488 ? Sl 10:23 3:21 /usr/sbin/zabbix_proxy: snmp poller #6 [got 83 values, queued 82 in 5 sec, awaiting 49] 1997 62317 25.6 0.2 1250020 30520 ? Sl 10:23 3:21 /usr/sbin/zabbix_proxy: snmp poller #6 [got 83 values, queued 82 in 5 sec, awaiting 49] 1997 62317 25.6 0.2 1276980 30884 ? Rl 10:23 3:21 /usr/sbin/zabbix_proxy: snmp poller #6 [got 83 values, queued 82 in 5 sec, awaiting 49] 1997 62317 25.6 0.2 1277048 30820 ? Rl 10:23 3:22 /usr/sbin/zabbix_proxy: snmp poller #6 [got 80 values, queued 70 in 5 sec, awaiting 39] 1997 62317 25.6 0.2 1250012 30592 ? Sl 10:23 3:22 /usr/sbin/zabbix_proxy: snmp poller #6 [got 80 values, queued 70 in 5 sec, awaiting 39]
but other SNMP pollers no longer pick up values:
$ while : ; do sleep 1; ps auxw|grep -i snmp|awk '$16 > 0 && $14 != "#6" {print $14}'; done
^ the output is empty, so no other poller is displayed at this moment.
queue after restart is growing after a while:
