Folks,

We had MAJOR problems using exportfs -f especially during a name service outage. (major = cratered a filer)

Working with escalations at NA we found that there could be a 45 minute wait to have new root nodes show up in your netgroup. We also were told that there was no EASY way to trigger a refresh of the netgroup data. We were, however, told NOT to do an exportfs -f (which had been our standard method). Apparently I was not alone in the way I was dealing with this.

The bug and RFE for an easier way to update the netgroup for root=@netgroup is logged but not public yet.

I have a script that does several steps to force the netgroup update but I want to clear this with my escalations engineer before I publish it.

I will send out an update to this email after working with him. C-

We have found that if you edit a local /etc/netgroup file on the filer to add a new NFS client, then it takes a while for the new client be be able to mount. If you don't want to wait, then you can re-export what the new client wants to mount or you can run

exportfs -f

to flush the export cache.

Steve Losen scl@virginia.edu phone: 434-924-0640

University of Virginia ITC Unix Support