+-- Anthony Fiarito alf@cp.net once said: | We are getting a bunch of: | | Tue Mar 30 17:59:13 GMT [sm_recover]: no address for host | <host.somedomain.com> | | messages on a few of our filers. Anyone know of a (safe) way to get | rid of them? The hostname and IP do not exist in /etc/hosts or | /etc/exports. Would removing /etc/sm/* get rid of these messages?

Yes. We've had to do this too. Here's the FAQ on this from the NetApp NOW site:

Q: What do the following messages mean?

[sm_recover]: no address for host <my_pc> [sm_recover]: get RPC port for <host=linux2,prog=100024,ver=1,prot=17> failed

A: In the first case ("no address for host"), the Filer is sending a status message to the client, following a filer reboot. The client specified in this message is perhaps 1) no longer on the net, 2) has had its IP address changed, or 3) as is the case with most PC clients, is not running rpc.statd. In the second case, sm_recover messages are issued because the server attempts to reestablish locks with client(s).

In both cases, these messages can be stopped by deleting all of the files in the /etc/sm directory on the filer; but do not delete the /etc/sm/ directory itself. The files can be removed and at any time and will be regenerated.

Oz