Linux Centos 8.5
Sprint 85 (Feb 2022), Sprint 86 (Mar 2022), Sprint 87 (Apr 2022), Sprint 88 (May 2022), Sprint 89 (Jun 2022), Sprint 90 (Jul 2022), Sprint 91 (Aug 2022), Sprint 92 (Sep 2022), Sprint 93 (Oct 2022), Sprint 94 (Nov 2022)
Steps to reproduce:
- System worsk initially as expected and monitors MongoDB
- MongoDB hangs and stop responding
- Zabbix Agent2 zabbix_get calls to MongoDB plugin related Item keys hangs too
- Other zabbix_get calls like agent.ping works well
Example (IP address and password obfuscated)
- time zabbix_get -s 0.0.0.0 -k mongodb.ping["tcp://localhost:00000","zab","xyz"]
zabbix_get : Timeout while executing operation
Zabbix_get call timeouts after1 minute - network timeout.
The consequence is, that Zabbix marks whole interface on the monitored Host as unsupported because of the network timeout and after while Zabbix marks the Interface working again because other Items like agent.ping works.
This flapping causes huge load by "unsupported pollers".
Expected result is that Zabbix agent returns Item value or returns ZBX_NOTSUPPORTED after agent timeout (3 secs) is reached.