1rpc.statd(8) System Manager's Manual rpc.statd(8)
2
6 rpc.statd - NSM status monitor
7
9 rpc.statd [-FNL] [-d] [-?] [-n name ] [-o port ] [-p port ] [-H
10 prog ] [-V]
11
13 The rpc.statd server implements the NSM (Network Status Monitor) RPC
14 protocol. This service is somewhat misnamed, since it doesn't actually
15 provide active monitoring as one might suspect; instead, NSM implements
16 a reboot notification service. It is used by the NFS file locking ser‐
17 vice, rpc.lockd, to implement lock recovery when the NFS server machine
18 crashes and reboots.
19 Operation
21 For each NFS client or server machine to be monitored, rpc.statd cre‐
22 ates a file in /var/lib/nfs/statd/sm. When starting, it normally runs
23 sm-notify to iterate through these files and notify the peer rpc.statd
24 on those machines.
25
27 -F By default, rpc.statd forks and puts itself in the background
28 when started. The -F argument tells it to remain in the fore‐
29 ground. This option is mainly for debugging purposes.
30 -d By default, rpc.statd sends logging messages via syslog(3) to
32 system log. The -d argument forces it to log verbose output to
33 stderr instead. This option is mainly for debugging purposes,
34 and may only be used in conjunction with the -F parameter.
35 -n, --name name
37 specify a name for rpc.statd to use as the local hostname. By
38 default, rpc.statd will call gethostname(2) to get the local
39 hostname. Specifying a local hostname may be useful for machines
40 with more than one interfaces.
41 -o, --outgoing-port port
43 specify a port for rpc.statd to send outgoing status requests
44 from. By default, rpc.statd will ask portmap(8) to assign it a
45 port number. As of this writing, there is not a standard port
46 number that portmap always or usually assigns. Specifying a
47 port may be useful when implementing a firewall.
48 -p, --port port
50 specify a port for rpc.statd to listen on. By default,
51 rpc.statd will ask portmap(8) to assign it a port number. As of
52 this writing, there is not a standard port number that portmap
53 always or usually assigns. Specifying a port may be useful when
54 implementing a firewall.
55 -P, --state-directory-path directory
57 specify a directory in which to place statd state information.
58 If this option is not specified the default of /var/lib/nfs is
59 used.
60 -N Causes statd to run in the notify-only mode. When started in
62 this mode, the statd program will check its state directory,
63 send notifications to any monitored nodes, and exit once the
64 notifications have been sent. This mode is used to enable Highly
65 Available NFS implementations (i.e. HA-NFS). This mode is dep‐
66 recated - sm-notify should be used directly instead.
67 -L, --no-notify
69 Inhibits the running of sm-notify. If sm-notify is run by some
70 other script at boot time, there is no need for statd to start
71 sm-notify itself. This can be appropriate if starting of statd
72 needs to be delayed until it is actually need. In such cases
73 sm-notify should still be run at boot time.
74 -H, --ha-callout prog
76 Specify a high availability callout program, which will receive
77 callouts for all client monitor and unmonitor requests. This
78 allows rpc.statd to be used in a High Availability NFS (HA-NFS)
79 environment. The program will be run with 3 arguments: The
80 first is either add-client or del-client depending on the reason
81 for the callout. The second will be the name of the client.
82 The third will be the name of the server as known to the client.
83 -? Causes rpc.statd to print out command-line help and exit.
85 -V Causes rpc.statd to print out version information and exit.
87
92 This rpc.statd version is protected by the tcp_wrapper library. You
93 have to give the clients access to rpc.statd if they should be allowed
94 to use it. To allow connects from clients of the .bar.com domain you
95 could use the following line in /etc/hosts.allow:
96 statd: .bar.com
98 You have to use the daemon name statd for the daemon name (even if the
100 binary has a different name).
101 For further information please have a look at the tcpd(8) and
103 hosts_access(5) manual pages.
104
107 SIGUSR1 causes rpc.statd to re-read the notify list from disk and send
108 notifications to clients. This can be used in High Availability NFS
109 (HA-NFS) environments to notify clients to reacquire file locks upon
110 takeover of an NFS export from another server.
111
114 /var/lib/nfs/statd/sm/state
115 /var/lib/nfs/statd/sm/*
116 /var/lib/nfs/statd/sm.bak/*
117
119 rpc.nfsd(8), portmap(8)
120
122 Jeff Uphoff <juphoff@users.sourceforge.net>
123 Olaf Kirch <okir@monad.swb.de>
124 H.J. Lu <hjl@gnu.org>
125 Lon Hohberger <hohberger@missioncriticallinux.com>
126 Paul Clements <paul.clements@steeleye.com>
127 31 Aug 2004 rpc.statd(8)