SNMP Service on Client stops working after approximate 24h

[Daniel4711]

Hi all, I have a Linux spam filter here where the SNMP service crashes at regular intervals (about 1x per day).
In my opinion there are not too many items and the polling interval is not too short.
1x 1min
22x 2min
2x 3min
32x 5min
8x 10min
1x 1h

I have already corrected the time interval of the items and also already performed tests with the bulk queries.
The SNMP service is no longer accessible after about a day.
The service must then be restarted on the appliance.

We don't have root access on the appliance and I hope someone else has a hint for me and the appliance support.

Details: Zabbix 5.4, SNMP 2c

[Daniel4711]

Of course I have contacted the support.
But they don't seem to have any further ideas either.
If you do some research, a crash of the Linux SNMP service is not uncommon when it is monitored or queried by the monitoring system.
Mostly without a solution.
I will ask the support again.

[Daniel4711]

Hello again,

the Support would like to know if Zabbix makes the queries via snmpget or snmpwalk.
I could not find any info here unfortunately.
If I see it correctly, zabbix uses the SNMP functions of the Linux system (Debian 10 for me).
I could not find a logfile here.

[markfree]

Hello again,

the Support would like to know if Zabbix makes the queries via snmpget or snmpwalk.
I could not find any info here unfortunately.
If I see it correctly, zabbix uses the SNMP functions of the Linux system (Debian 10 for me).
I could not find a logfile here.

Zabbix uses either SNMP GetRequest and GetBulkRequest, so, it uses SNMP Get and Walk.

SNMP documentation describes

technical issue that not all devices are capable of returning 128 values per request

You could check your host SNMP agent capabilities.

Zabbix also offer a setting:

since Zabbix 2.4 there is a “Use bulk requests” setting for each interface that allows to disable bulk requests for that device.

You can verify SNMP Internal workings here.

[cyber]

That spam filter thingie is an appliance? And you have support for it... why does support not resolve the issue of that snmp service crashes?
From here it does not really look like a Zabbix issue......