I had an NFS server motherboard go bad, and a quick replacement resulted in the drives coming up in all the wrong places, and no room to do a new install. I'm running at the moment off a USB boot to keep thinks going until I get an SSD so I can install there. A thumb drive on a USB 3.0 port is not a bad substitute for an SSD short term.

Anyway, the issue is ID mapping from the server to the clients. The documentation is written using terms like
  "NFS4_user_object@REALM = local_user_object"
and while I have the more recent clients working, machines lacking the rpc.idmap (aka nfs-idmap) capability aren't working. Since I have to support the existing clients rather than having the ability to force sensible upgrades, and many of the clients just don't do idmap, two questions.

FIRST - is there a human readable document with examples which just show what client and server mods to /etc/idmapd.conf should be made such that j...@foo.demo.org gets access to the joe files on the server, and what the effect of those changes is are on the client and server sides. That's sort of medium critical, since I have that working. It's unclear if and how some of the terms map to OO languages.

SECOND - is there a way to get the server to provide NFS3 service for the machines which were put in service before the even was an NFS4? Other than running a whole different server? I see that it is supposed to happen if /proc/fs nsf* are unmounted, but that creates breakage. Is there something simple to make that happen?

All I need is some doc so I can be sure I've done things right, as opposed to hacked until they work.
--
Bill Davidsen <david...@tmr.com>
  "We have more to fear from the bungling of the incompetent than from
the machinations of the wicked."  - from Slashdot
--
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Reply via email to