SNMP Version 5.5
Spine 1.1.36
378 devices, a mix of Windows servers (SNMPv2) and network switches (SNMPv2/v3)
cmd polling was working, but as we have added devices bit by bit it is now starting to not complete within 5 minutes. So I thought try switching to the Spine poller.
The issue is weirdly every poll run it seems to keep thinking various devices are DOWN, it seems a bit random as to the devices picked too. When you go into them manually in Cacti the SNMP summary and ping response at the top is fine (ping is set for ICMP and responds well under the 500ms timeout)
Spine.exe when run manually also comes up with a lot of 'use snmp_sess_select_info2() for processing large file descriptors' errors
So wondering where I am going wrong as this would otherwise be a much faster poller!
