I am dealing with a lot of SNMP requests (snmp flood) to Cisco devices after device was unavailable to Zabbix for some time.
Problem started after upgrading from zabbix 5 to zabbix 6.4 in docker environment.
After connection to the cisco device is restored, Zabbix keeps sending a lot of SNMP requests to the device and CPU on that Cisco device jumps from 15% to over 80%.
In Zabbix 6.4 by default on every HOST SNMP interface has "Max repetition count" set to 10 and checked "Use combined requests".
When that happens, reducing Max repetition and unchecking "Use combined request" does nothing at all. Zabbix still sends a lot of requests.
Even if host is disabled, Zabbix is still sending SNMP requests. And no, there is no other host with an identical interface.
Restarting the zabbix container solves problem, but i dont want to do that every time after some device becomes unavailable.
Is that a bug or a feature or am I doing something wrong?
Problem started after upgrading from zabbix 5 to zabbix 6.4 in docker environment.
After connection to the cisco device is restored, Zabbix keeps sending a lot of SNMP requests to the device and CPU on that Cisco device jumps from 15% to over 80%.
In Zabbix 6.4 by default on every HOST SNMP interface has "Max repetition count" set to 10 and checked "Use combined requests".
When that happens, reducing Max repetition and unchecking "Use combined request" does nothing at all. Zabbix still sends a lot of requests.
Even if host is disabled, Zabbix is still sending SNMP requests. And no, there is no other host with an identical interface.
Restarting the zabbix container solves problem, but i dont want to do that every time after some device becomes unavailable.
Is that a bug or a feature or am I doing something wrong?
Attached Files