1OCF_HEARTBEAT_FILESY(7) OCF resource agents OCF_HEARTBEAT_FILESY(7)
2
6 ocf_heartbeat_Filesystem - Manages filesystem mounts
7
9 Filesystem [start | stop | monitor | meta-data | validate-all]
10
12 Resource script for Filesystem. It manages a Filesystem on a shared
13 storage medium.
14 The standard monitor operation of depth 0 (also known as probe) checks
16 if the filesystem is mounted. If you want deeper tests, set
17 OCF_CHECK_LEVEL to one of the following values:
18 10: read first 16 blocks of the device (raw read)
20 This doesn't exercise the filesystem at all, but the device on which
22 the filesystem lives. This is noop for non-block devices such as NFS,
23 SMBFS, or bind mounts.
24 20: test if a status file can be written and read
26 The status file must be writable by root. This is not always the case
28 with an NFS mount, as NFS exports usually have the "root_squash" option
29 set. In such a setup, you must either use read-only monitoring
30 (depth=10), export with "no_root_squash" on your NFS server, or grant
31 world write permissions on the directory where the status file is to be
32 placed.
33
35 device
36 The name of block device for the filesystem, or -U, -L options for
37 mount, or NFS mount specification.
38 NOTE: On Linux /dev/disk/by-{uuid,label}/ are preferred to -U/-L.
40 (required, string, no default)
42 directory
44 The mount point for the filesystem.
45 (required, string, no default)
47 fstype
49 The type of filesystem to be mounted.
50 (required, string, no default)
52 options
54 Any extra options to be given as -o options to mount.
55 For bind mounts, add "bind" here and set fstype to "none". We will
57 do the right thing for options such as "bind,ro".
58 (optional, string, no default)
60 statusfile_prefix
62 The prefix to be used for a status file for resource monitoring
63 with depth 20. If you don't specify this parameter, all status
64 files will be created in a separate directory.
65 (optional, string, default ".Filesystem_status/")
67 run_fsck
69 Specify how to decide whether to run fsck or not.
70 "auto" : decide to run fsck depending on the fstype(default)
72 "force" : always run fsck regardless of the fstype "no" : do not
73 run fsck ever.
74 (optional, string, default "auto")
76 fast_stop
78 Normally, we expect no users of the filesystem and the stop
79 operation to finish quickly. If you cannot control the filesystem
80 users easily and want to prevent the stop action from failing, then
81 set this parameter to "no" and add an appropriate timeout for the
82 stop operation.
83 This defaults to "no" for GFS2 filesystems.
85 (optional, boolean, default no)
87 force_clones
89 The use of a clone setup for local filesystems is forbidden by
90 default. For special setups like glusterfs, cloning a mount of a
91 local device with a filesystem like ext4 or xfs independently on
92 several nodes is a valid use case.
93 Only set this to "true" if you know what you are doing!
95 (optional, boolean, default false)
97 force_unmount
99 This option allows specifying how to handle processes that are
100 currently accessing the mount directory.
101 "true" : Kill processes accessing mount point "safe" : Kill
103 processes accessing mount point using methods that avoid functions
104 that could potentially block during process detection "false" : Do
105 not kill any processes.
106 The 'safe' option uses shell logic to walk the /procs/ directory
108 for pids using the mount point while the default option uses the
109 fuser cli tool. fuser is known to perform operations that can
110 potentially block if unresponsive nfs mounts are in use on the
111 system.
112 (optional, string, default "safe")
114 term_signals
116 Signals (names or numbers, whitespace separated) to send processes
117 during graceful termination phase in stop-action.
118 (optional, string, default "TERM")
120 kill_signals
122 Signals (names or numbers, whitespace separated) to send processes
123 during forceful killing phase in stop-action.
124 (optional, string, default "KILL")
126 signal_delay
128 How many seconds to wait after sending term/kill signals to
129 processes in stop-action.
130 (optional, string, default "1")
132
134 This resource agent supports the following actions (operations):
135 start
137 Starts the resource. Suggested minimum timeout: 60s.
138 stop
140 Stops the resource. Suggested minimum timeout: 60s.
141 monitor
143 Performs a detailed status check. Suggested minimum timeout: 40s.
144 Suggested interval: 20s.
145 validate-all
147 Performs a validation of the resource configuration. Suggested
148 minimum timeout: 5s.
149 meta-data
151 Retrieves resource agent metadata (internal use only). Suggested
152 minimum timeout: 5s.
153
155 The following is an example configuration for a Filesystem resource
156 using the crm(8) shell:
157 primitive p_Filesystem ocf:heartbeat:Filesystem \
159 params \
160 device=string \
161 directory=string \
162 fstype=string \
163 op monitor depth="0" timeout="40s" interval="20s"
164
166 The following is an example configuration for a Filesystem resource
167 using pcs(8)
168 pcs resource create p_Filesystem ocf:heartbeat:Filesystem \
170 device=string \
171 directory=string \
172 fstype=string \
173 op monitor OCF_CHECK_LEVEL="0" timeout="40s" interval="20s"
174
176 http://clusterlabs.org/
177
179 ClusterLabs contributors (see the resource agent source for information
180 about individual authors)
181resource-agents UNKNOWN 10/11/2023 OCF_HEARTBEAT_FILESY(7)