Linux/Solaris nfs and ACLs

From: Mark Andrews (marka@calibredigital.com)
Date: Wed Feb 11 2004 - 08:41:00 EST


I've got two issues that i believe are related.

We have two Sunfire V480Rs (Solaris 9, 64-bit SPARC) in a veritas
cluster (VCS 3.5p2) in an HA SAN configuration. Both sunfires are
connected to a Storedge 6320 via 2Gb fibre channel. The array has
two 2TB vxfs cluster filesystems called array1 and array2 mounted
on each of the sunfires by the cluster software and nfs shared to
a few samba servers and nfs clients.

All of the nfs clients are IRIX or RedHat Linux 7.3/9.0 with
a 2.4 or 2.6 kernel. The cluster software has shared the two
filesystems with the -root=<fully qualified name> option:

/export/array1 rw,anon=60001,root=FQDN1:FQDN2:FQDN3
/export/array2 rw,anon=60001,root=FQDN1:FQDN2:FQDN3

yet when i write to the filesystem from an NFS client as the
superuser, the uid is remapped to the anon uid.

Am i missing something or is there a trick to this?

A separate issue involves ACLs on the same filesystems. From the
nfs clients, if i run "getfacl file" on a nfs file or directory,
i only see the default unix permissions and no ACL information.

If i attempt to modify the ACL of a file or directory, i get
an "Operation not supported" error.

I notice that there are a few RPC services on the sun nfs servers
called nfs_acl, but don't know how to specifically access them.

Any assistance would be appreciated.

Mark
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers



This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 23:28:02 EDT