HP Procurve Manager

[andy_johnson(cnews)]

Hello Everyone!!

I have recently installed hp procurve manager so that we have more control over our core network. Unfortunately it is only picking up about 25% of our switches? I have looked at all device configs and I can not see any difference in the way snmp is set on the devices that are not being discovered by hp procurve manager. (Some expert help needed i think!!!!)
I would be happy to provide more information & running configs if needed.

Thanks in advance,

Andy Johnson

[cenk sasmaztin]

hi Andy

please read this guide

http://cdn.procurve.com/training/Manuals/PCM23-AdminGuide59908850-1207.pdf

[Javed Padinhakara]

Some troubleshooting tips that might help:

- Is there any pattern in the set of devices that were not discovered? - for example all of them are of some different subnet?
- Check if any of these devices have been for some reason listed under "Unknown Devices" in the Device Tree on left display in PCM.
- Check if the configurations in Tools>Preferences>Discovery ( like exclude devices, managed subnet etc) are all OK.
- Launch Tools> Manual Discovery Wizard and specify IP of a ( carefully determined and not yet discovered) device through which many others in the network can possibly be reached. Uncheck "Use defaults " and supply correct SNMP/CLI credentials for that device.
- If the above operation fails, check few logs under server\logs\ like Disco*.log, Cs-err*.log, Cs-out*.log and see for some clue - there could be some messages giving an hint to the problem.

Hope that helps
Javed

[andy_johnson(cnews)]

Thanks for you for getting back to me

I can see all switches that have not been recognised in 'unknown devices'.
It appears that the switches in unknown devices all share the same ip range '10.0.35.x(edges) apart from one 10.0.2.4(second core ).
I have added the 10.0.0.0 & 10.0.35.0 to pcm global discovery but its still not finding anything?
When I run manual discovery I get:

Checking read access community name...
Read community name failed, operation timed out

Checking write access community name...
Write community name failed, operation timed out

Unable to communicate with device 10.0.35.28

Another thing is our core switches are 10.0.2.3 & 10.0.2.4 pcm is happy to pick up 10.0.2.3 but is blissfully ignores 10.0.2.4.

Thanks in advance

Andy Johnson

[andy_johnson(cnews)]

Ok I think I have worked this one out it seems that the time out is being caused by the snmp traffic having no route back from vlan 135(switches) to vlan 100(servers).

I will assigning an ip address to vlan 100 and applying a static routes help?

[Javed Padinhakara · Accepted solution]

Yes Andy. That should do the trick.

The timeout error is usually because of unavailability of the PCM server to communicate with the the device.

You can setup the routes such that from PCM server the devices are pingable atleast.

Do let know how that goes about

[andy_johnson(cnews)]

Thanks for your advice I have done the configuraion that was necessary and can now see all our switches using pcm.

Thanks Again

Andy Johnson

[willinfo]

Hi Andy. I have the same problem with SNMP and community name. I'm interested to know how you resolved your problem .
Thanks you to respond.

[andy_johnson(cnews)]

The issue I experienced was caused by our vlan's set up! As it turned out snmp was timing out because it could not find the ip address I supplied to the manual discovery wizard.

My problem was solved by double and triple checking the snmp settings and then assigning an ip address to the corresponding vlan so that snmp has connectivity.

Cheers

Andy

[willinfo]

It's possible that you tell me exactly what you did because I don't know how to do ?