1clurmtabd(8)                 Red Hat Cluster Suite                clurmtabd(8)
2
3
4

NAME

6       clurmtabd - Cluster NFS Remote Mount Table Daemon
7

DESCRIPTION

9       Instances  of clurmtabd keep entries relevant to a given mount point in
10       sync cluster-wide.  It does this by polling /var/lib/nfs/rmtab periodi‐
11       cally.
12

HOW IT WORKS

14       A  single  instance  of  clurmtabd  is  spawned  by the service manager
15       (clusvcmgrd(8)) for each mount point in a  given  user  service.   Each
16       instance  polls  /var/lib/nfs/rmtab  every  few  seconds,  looking  for
17       changes.  When changes are detected, they are merged in with  the  cur‐
18       rent  in-memory  view.  The entries are filtered according to the mount
19       point being monitored and are subsequently written out synchronously to
20       that mount point.
21
22       When  a  user  service  is  relocated (or after a fail-over), clurmtabd
23       reads the entries on a given mount point and synchronizes them with the
24       new  member's  /var/lib/nfs/rmtab.   This  is done prior to running the
25       exportfs(8) command so that clients of the given service do not receive
26       "Stale NFS file handle" errors.
27

NOTES

29       Clurmtabd  is not a part of NFS and can not ensure that "Stale NFS file
30       handle" errors are never received.  It is a  best-effort  attempt.   To
31       fully prevent "Stale NFS file handle" errors, it is necessary to modify
32       the Linux kernel as well as nfs-utils.
33
34

FILES

36       /var/lib/nfs/rmtab
37
38

SEE ALSO

40       clurgmgrd(8), rpc.mountd(8), exportfs(8)
41
42
43
44                                   Jan 2005                       clurmtabd(8)
Impressum