topic Re: ARP settings and problems in OpenVMS in Operating System - OpenVMS

ARP settings and problems in OpenVMS

Piet Timmers_1 — Mon, 20 Dec 2010 08:42:04 GMT

We have three OpenVMS nodes, TCPIP version OpenVMS Alpha Version V5.0A, in one location, and on one switch.
The network provider has changed some settings for the network. Two nodes cannot be reached, one is normal.

The network provider has disabled ARP-PROXY.
What can we do to get everything working, whithout using ARP-PROXY?

Thanks.

Piet

Re: ARP settings and problems in OpenVMS

Oswald Knoppers_1 — Mon, 20 Dec 2010 10:10:53 GMT

I am not sure i understand your network configuration.

Are all three nodes connected to the same switch? If so, i fail to see what arp-proxy has to do with it.

Oswald

Re: ARP settings and problems in OpenVMS

Piet Timmers_1 — Mon, 20 Dec 2010 10:28:15 GMT

Oswald,

Thats also my problem.

They disabled ARP-PROXY on a router and we had a problem reaching two out of three nodes. A windows domaincontroller, on the same switch, was reachable.

They enabled ARP-PROXY, no problem.

And yes, the three nodes are connected to the same switch.

Greetings

Re: ARP settings and problems in OpenVMS

Oswald Knoppers_1 — Mon, 20 Dec 2010 10:37:39 GMT

Are all three nodes in the same (sub)net?
Can you do a '$ tcpip show arp' on all three nodes?

Oswald

Re: ARP settings and problems in OpenVMS

Richard Brodie_1 — Mon, 20 Dec 2010 10:49:54 GMT

One of the things an ARP proxy can do, is to make up for a bad subnet configuration.

I suggest you review your network settings carefully, particularly the subnet mask.

Alternatively, you could take some packet captures (e.g. with Wireshark) and work out what's different. That's likely to be harder work though.

Re: ARP settings and problems in OpenVMS

Piet Timmers_1 — Mon, 20 Dec 2010 10:53:02 GMT

The output formatting is better in an external document. See attachment.

Re: ARP settings and problems in OpenVMS

Richard Brodie_1 — Mon, 20 Dec 2010 11:05:06 GMT

This was a very recent change, then?

It looks like the unreachable nodes have stale proxy ARP entries:

e.g. 10.114.37.237 00-00-0c-07-ac-01

whereas the working node has an ARP entry for the gateway.

10.114.3.1 00-00-0c-07-ac-01

I think it might be OK when the stale entries timeout. A few "SET NOARP ip", to flush them, should speed things up.

Re: ARP settings and problems in OpenVMS

STEPHEN CONWAY_1 — Thu, 30 Dec 2010 23:03:14 GMT

The problem is not that 'Two nodes cannot be reached', the problem is that two nodes (ASDS9C and PAML01) do not see their default gateway (10.114.3.1). It is not directly connected.