gfs_controld(8)

1GFS_CONTROLD(8)                     cluster                    GFS_CONTROLD(8)
2
3
4

NAME

6       gfs_controld - daemon that coordinates gfs mounts and recovery
7
8

SYNOPSIS

10       gfs_controld [OPTIONS]
11

DESCRIPTION

13       GFS  lives in the kernel, and the cluster infrastructure (corosync mem‐
14       bership and group management) lives in user space.  GFS in  the  kernel
15       needs  to  adjust/recover  for certain cluster events.  It's the job of
16       gfs_controld to receive these events and  reconfigure  gfs  as  needed.
17       gfs_controld  controls  and configures gfs through sysfs files that are
18       considered gfs-internal interfaces.
19
20       Mounting and node failure are the main cluster events that gfs_controld
21       controls.   It  also  manages  the  assignment of journals to different
22       nodes.  mount.gfs2(8) communicates with gfs_controld to join the  mount
23       group and receive the necessary options for the kernel mount.
24
25       The cman init script usually starts the gfs_controld daemon.
26
27

OPTIONS

29       Command line options override a corresponding setting in cluster.conf.
30
31
32       -D     Enable debugging to stderr and don't fork.
33              See also gfs_control dump in gfs_control(8).
34
35
36       -L     Enable debugging to log file.
37              See also logging in cluster.conf(5).
38
39
40       -g num groupd compatibility mode, 0 off, 1 on.
41              Default 0.
42
43
44       -w num Enable (1) or disable (0) withdraw.
45              Default 1.
46
47
48       -h     Print a help message describing available options, then exit.
49
50
51       -V     Print program version information, then exit.
52
53
54   Cluster Upgrading
55       When  performing  a  rolling  upgrade  from  cluster2  to cluster3, the
56       gfs_controld daemon runs in cluster2 compatibility mode  and  processes
57       posix  lock  (plock)  requests  from  the dlm to remain compatible with
58       cluster2 nodes.  Otherwise, dlm_controld(8) handles plocks.
59
60
61       -p num Enable (1) or disable (0) plock code for cluster fs.
62              Default 1.
63
64
65       -l num Limit the rate of plock operations, 0 for no limit.
66              Default 0.
67
68
69       -o num Enable (1) or disable (0) plock ownership.
70              Default 0.
71
72
73       -t ms  Plock ownership drop resources time (milliseconds).
74              Default 10000.
75
76
77       -c num Plock ownership drop resources count.
78              Default 10.
79
80
81       -a ms  Plock ownership drop resources age (milliseconds).
82              Default 10000.
83
84
85       -P     Enable plock debugging messages (can produce excessive output).
86
87

FILES

89       cluster.conf(5) is usually located at /etc/cluster/cluster.conf.  It is
90       not  read  directly.   Other  cluster components load the contents into
91       memory, and the values are accessed through the libccs library.
92
93       Configuration options are added to  the  <gfs_controld  />  section  of
94       cluster.conf, within the top level <cluster> section.
95
96
97       enable_withdraw
98              Set to 1/0 to enable/disable a response to a withdraw.
99
100              <gfs_controld enable_withdraw="1"/>
101
102
103   Cluster Upgrading
104       See  note  above  about  compatibility  plock handling during a rolling
105       upgrade.  The following options are deprecated, see dlm_controld(8) for
106       the new equivalent options.
107
108
109       enable_plock
110              See command line description.
111
112              <gfs_controld enable_plock="1"/>
113
114
115       plock_rate_limit
116              See command line description.
117
118              <gfs_controld plock_rate_limit="0"/>
119
120
121       plock_ownership
122              See command line description.
123
124              <gfs_controld plock_ownership="0"/>
125
126
127       drop_resources_time
128              See command line description.
129
130              <gfs_controld drop_resources_time="10000"/>
131
132
133       drop_resources_count
134              See command line description.
135
136              <gfs_controld drop_resources_count="10"/>
137
138
139       drop_resources_age
140              See command line description.
141
142              <gfs_controld drop_resources_age="10000"/>
143
144
145       plock_debug
146              Enable  (1) or disable (0) plock debugging messages (can produce
147              excessive output). Default 0.
148
149              <gfs_controld plock_debug="0"/>
150
151

NAME

SYNOPSIS

DESCRIPTION

OPTIONS

FILES

SEE ALSO