[solved] Device Status Down - Graphs still populating

[cahonis]

Hi All,

Cacti is reporting several devices with a status of down.

When I go into the device it displays the SNMP system information correctly and they all have up to date graphs and data files.

I can not find any entries in the log when I filter by the host ID.

Windows Server 2003 SP1
Cacti 8.0.7g
PIA 2.9

poller is spine running every 5 minutes with the following settings:
Maximum Concurrent Poller Processes: 4
Balance Process Load: Yes
Maximum Threads per Process: 12
Number of PHP Script Servers: 4
Script and Script Server Timeout Value: 60
The Maximum SNMP OID's Per SNMP Get Request: 10

I've searched on here for a solution but cannot find one.

I recently changed from CMD to spine due to the poller running longer than 5 minutes.

I have run the Rebuild Poller Cache task.

Any assistance much appreciated.

[BSOD2600]

Have you changed the cacti logging level to high to find out why Spine thinks the device(s) are down? Spine 0.8.8a also fixes several issues with pings/host checking - http://www.cacti.net/spine_changelog.php

[cahonis]

Have you changed the cacti logging level to high to find out why Spine thinks the device(s) are down?

I hadn't but now have:

Code: Select all

06/26/2012 08:00:38 AM - SPINE: Poller[0] Host[492] DS[10763] SNMP: v2: DR-AV01, dsname: traffic_in, oid: .1.3.6.1.2.1.2.2.1.10.10, value: 1995936254
06/26/2012 08:00:36 AM - SPINE: Poller[0] Host[492] DS[10758] SS[1] SERVER: D:\cacti\scripts\ss_host_disk.php ss_host_disk DR-AV01 492 2:161:5000:2:1:closepfpub:::::: get total 2, output: 42842714112
06/26/2012 08:00:35 AM - SPINE: Poller[0] Host[492] DS[10763] SNMP: v2: DR-AV01, dsname: traffic_out, oid: .1.3.6.1.2.1.2.2.1.16.10, value: 813235839
012 08:00:35 AM - SPINE: Poller[0] Host[492] DS[10762] SS[1] SERVER: D:\cacti\scripts\ss_host_cpu.php ss_host_cpu DR-AV01 492 2:161:5000:2:1:closepfpub:::::: get usage 1, output: 2
06/26/2012 08:00:34 AM - SPINE: Poller[0] Host[492] DS[10761] SS[0] SERVER: D:\cacti\scripts\ss_host_cpu.php ss_host_cpu DR-AV01 492 2:161:5000:2:1:closepfpub:::::: get usage 0, output: 2
06/26/2012 08:00:34 AM - SPINE: Poller[0] Host[492] DS[10760] SS[3] SERVER: D:\cacti\scripts\ss_host_disk.php ss_host_disk DR-AV01 492 2:161:5000:2:1:closepfpub:::::: get used 5, output: 3693281280
06/26/2012 08:00:32 AM - SPINE: Poller[0] Host[492] DS[10760] SS[2] SERVER: D:\cacti\scripts\ss_host_disk.php ss_host_disk DR-AV01 492 2:161:5000:2:1:closepfpub:::::: get total 5, output: 4294443008
06/26/2012 08:00:32 AM - SPINE: Poller[0] Host[492] DS[10759] SS[0] SERVER: D:\cacti\scripts\ss_host_disk.php ss_host_disk DR-AV01 492 2:161:5000:2:1:closepfpub:::::: get used 4, output: 2601582592
06/26/2012 08:00:32 AM - SPINE: Poller[0] Host[492] DS[10759] SS[2] SERVER: D:\cacti\scripts\ss_host_disk.php ss_host_disk DR-AV01 492 2:161:5000:2:1:closepfpub:::::: get total 4, output: 8587051008
06/26/2012 08:00:31 AM - SPINE: Poller[0] Host[492] DS[10758] SS[3] SERVER: D:\cacti\scripts\ss_host_disk.php ss_host_disk DR-AV01 492 2:161:5000:2:1:closepfpub:::::: get used 2, output: 36546101248
06/26/2012 08:00:30 AM - SPINE: Poller[0] Host[492] PING Result: Ping not performed due to setting.

I don't understand why the last line reports "Ping not performed due to setting". Downed device detection is set to Ping or SNMP with ping method using ICMP.

I changed the ping method to SNMP only and the device is now in a recovery state.

You mentioned there are some ping/host checking bug fixes in 0.8.8a - would this be one of them? The changelog for 0.8.7i mentions a rework of ping but isn't specific.

[BSOD2600]

It's hard to tell since it doesn't appear the ping issue was associated to a specific bug, which makes it hard to investigate into what change(s) were made. TheWitness would be the resident expert, but has been MIA for a while due to other things. Best I can suggest is trying to newer versions of spine and see if that resolves your problem.... or change your downtime device detection method.

[cahonis]

Thanks for your replies, appreciated as always

I'll look at upgrading to the latest version and see if that helps.

The SNMP Get Mounted partitions script has now stopped working on new ESX hosts added since they had a VMWare patch which is weird, the existing ones are still polling information correctly.

I'm hoping an upgrade will fix that too

[BSOD2600]

The SNMP Get Mounted partitions script has now stopped working on new ESX hosts added since they had a VMWare patch which is weird, the existing ones are still polling information correctly.

The verbose output is also blank? Anything SNMP related from the host work?

[cahonis]

The SNMP Get Mounted partitions script has now stopped working on new ESX hosts added since they had a VMWare patch which is weird, the existing ones are still polling information correctly.

The verbose output is also blank? Anything SNMP related from the host work?

It would appear they do not like "The Maximum SNMP OID's Per SNMP Get Request" set to anything other than 1.

All working now.

Thanks again for you help