cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1011
Views
0
Helpful
8
Replies

HUM polling issue

gregory.aniorte
Level 1
Level 1

I've got a problem in HUM as I can't poll 5 devices. When I want to configure polling instances on those 5 switches (3500XL, 3550 and 3560 in snmp v2 and v3), there is nothing to poll.

snmpwalk is working fine and all other module are working fine on those devices.

It is doing that since I've restored LMS on my new server.

I'm using LM 3.1 with averything up to date.

I've got those errors in the upm_process.log

============================

Aug 25 15:16:01 EDT 2008 ],ERROR,[Thread-18],com.cisco.nm.upm.dal.DeviceAccess,initCredentials,130,SNMP credenttilas are not available with device manager.

Aug 25 15:15:19 EDT 2008 ],ERROR,[SNMPThrPool149],com.cisco.nm.upm.pollermgmt.instret.InstanceQueryRH,processSnmpPollTimeout,244,got the response processSnmpPollTimeout for device:10.16.0.22

[ Aug 25 15:15:19 EDT 2008 ],ERROR,[SNMPThrPool157],com.cisco.nm.upm.pollermgmt.instret.InstanceQueryRH,processSnmpPollTimeout,244,got the response processSnmpPollTimeout for device:10.16.0.25

[ Aug 25 15:15:19 EDT 2008 ],ERROR,[SNMPThrPool152],com.cisco.nm.upm.pollermgmt.instret.InstanceQueryRH,processSnmpPollTimeout,244,got the response processSnmpPollTimeout for device:10.16.0.29

[ Aug 25 15:15:19 EDT 2008 ],INFO ,[SNMPThrPool149],com.cisco.nm.upm.dal.DeviceAccess,changeToV1FromV2,816,SNMP Version is -1. So returning false.

[ Aug 25 15:15:19 EDT 2008 ],INFO ,[SNMPThrPool157],com.cisco.nm.upm.dal.DeviceAccess,changeToV1FromV2,816,SNMP Version is -1. So returning false.

[ Aug 25 15:15:19 EDT 2008 ],INFO ,[SNMPThrPool152],com.cisco.nm.upm.dal.DeviceAccess,changeToV1FromV2,816,SNMP Version is -1. So returning false.

====================================

Anyone ?

Thanks,

8 Replies 8

Joe Clarke
Cisco Employee
Cisco Employee

HUM will use the SNMP credentials in DCR just like every other module. From this log output, it appears those credentials are incorrect. You will not be able to use SNMPv3 on the 3500XL so make sure you do not have any v3 credentials in DCR for this device.

Start a packet capture filtering on udp/161 for one failing device, then try to setup a poller. Check the capture to see exactly what credential it is using to do the polling.

Ok, I've started a capture as you said.

I don't know how to distinguished the snmp request initiated by DFM from those initiated by HUM.

Nevertheless, I've got several snmp request and response that are using the good credential, I guess they are from DFM. But I've got some request that are using v2/public community, I guess there are from HUM.

It is the same for my 5 devices.

Is there a way to delete these devices from HUM and import them as we can do in DFM ?

Thanks.

Hi Joe,

Were you able to find something about my issue ?

My Ciscoworks server still sending v2/public community to devices configured in v3 in the DRC.

The problem exist since the restore.

Thanks.

HUM uses the same credentials that are in DCR. The public community string is used as a last ditch effort is a credential cannot be located. You can try deleting the pollers which contain these problem devices, then re-add them back. This can be done under HUM > Poller and Template Management > Poller Management.

Ok, i've deleted the pollers which cause problems. I've already done that before.

This time, I've stop/start the crmdmgtd service before reconfiguring the pollers.

It seems to work for my 5 devices, but the weird thing is I've go the same issue on other 2 devices. I've never get problem before that on those devices.

I'm using an evaluation version. The number of devices configured is 32 on the first poller (680 objects polled), and 3 on the second.

Is there another limitation, except the 100 devices limit, on the HUM eval licence ?

Thanks.

No, just the 100 device limit. You can check the upm_process.log to see if there are any errors pertaining to these two problem devices.

They are the same error, only IP addresses change.

During my last post, 2 other devices has caused the same problem.

I'm polling every 5 min, 700 objects polled every 5 min, isn't too much ?

I've got a big server (2 quad core, 8GB RAM, 400GB Raid5, and before the restauration there was no issue.

I'm using windows 2003 x64. Before the restauration I used windows 2003 32 bts.

Thanks.

This is fine. HUM can poll up to 60,000 objects every five minutes. While I haven't personally used Windows 64-bit, it is supported, and it should work.

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: