PCM+ Getting switches to update

[Sudain]

I'm unable to get my switches to update via PCM+. I can update my switches manually - but that invalidates the point of PCM. I scheudle the task, and theck the event log. I get 2 events:
* Download request Received from 172.16.0.10
* tftp: connection to 129.82.133.x failed

The PCM server is dual-homed with both 172.16.0.10 and 129.82.133.x addresses. I've added both addresses in the server .\pnm\server\config\commIpAddr.txt and .\pnm\client\config\commIpAddr.txt with 172.16.0.10 taking prescidence.

Ideas?

[AUS Adam]

my PCM app is also on a dual-homed server but i only put one ip address in the commIpAddr.txt file and that is of the nic that presents to the LAN

[Sudain]

Swapped that out and no go.

Something else I noticed that's odd is when I go into the switch and select downloading a tftp client it is set to 129.82.133.x.

When I did a port scan on the machine's normal IP I got the below ports none of which appear to be tftp related.

21/tcp
42/tcp
135/tcp
445/tcp
1027/tcp
1049/tcp
1068/tcp
3389/tcp
8009/tcp
9001/tcp
30000/tcp

[Javed Padinhakara]

Which version of PCM are you using?

[Sudain]

Version: PCM C.03.10.201
This is the free 2 month trial as well.

[Javed Padinhakara]

ok..so which subnet are your devices in - 172.x.x.x or 129.x.x.x ?

Here is one trick you can try:
a) Stop PCM server and Agent Services
b) Create a file "commIpAddr.txt" under "\pnm\pcm-agent\config" directory.
c) Add entry 172.16.0.10 in that file ( same as the one you added in .\pnm\server\config\commIpAddr.txt )
d) Restart PCM Server and Agent

Check using netstat or currports, to ensure both PCM Server and Agent are bound to same IP (amongst the dual IP's)

Now attempt your software update

Pl let know how this goes about

HTH
`Javed

[Sudain]

My clients and all end-user machines are on the 129.x.x.x scope - this is the IP I connect to via RDP. I installed PCM on this via this address - I only enabled the 172.x.x.x network connection after the installation was complete. I don't know if that's relevent or not. This is also my DHCP server for the 129.x.x.x subnet though I'm not sure that has anything to do with anything.


The 172.x.x.x is the scope that has only the switches - nothing else should ever be on that scope. This is also a non-routeable IP scope on a vlan that doesn't go to the router.

I wasn't sure what you meant by 'restart the agent' so I restarted the PCM services, deleted and re-discovered the device. with no increase in luck.

Using netstat I noticed that port 69 is bound to the 129.x.x.x address. Not sure that's relevent or not either.

UDP 172.16.0.10:67 *:*(Scvhost)
UDP 172.16.0.10:68 *:*(Scvhost)
UDP 129.82.133.x:67 *:*(Scvhost)
UDP 129.82.133.x:68 *:*(Scvhost)
UDP 129.82.133.x:69 *:*(javaw.exe*32) - PCM?

[KSimpson]

I had the same problem. We had our server dual -homed and it would never upgrade the firmware on the switches on one of the subnets but however it would scan the device. After opening a ticket and HP not able to help resolve the issue, I created a management vlan and added all our switches and the server to the network. Everything works fine now. I played around with that file many of days. It never helped. I was gettign the same log files as well. You might want to put all your managed switches on a management vlan. It will probably resolve your issue a lot quicker. It would will be more secure also.

[Sudain]

I like the way you are thinking... :)

However my 172.16.x.x is my managment vlan - all of my switches have Ips on that scope and I have verified connectivity through it by hand. The onlything it's missing is layer 3 services as I don't control a router.

Or am I misunderstanding what you are suggesting?

[KSimpson]

Well. If all your switches and your PCM Server are all on the same network, you shouldn't need layer 3 for them to communicate to each other. Just remove the 129 network from the server. How many interfaces (and/or IPs) do you have on the PCM server?