1CLUSTER.CONF(5) cluster CLUSTER.CONF(5)
2
3
4
6 cluster.conf - configuration file for cman and related daemons
7
8
10 /etc/cluster/cluster.conf
11
12
14 When cman_tool(8) starts the corosync(8) daemon, the cluster.conf data
15 is read into the corosync in-memory database (confdb). The configura‐
16 tion is used by corosync, cman and other related cluster daemons and
17 programs. When cman configures corosync with cluster.conf, the
18 corosync.conf(5) file is not used.
19
20 A basic cluster configuration is described below. Configuration
21 options for other daemons/programs are described in their own man
22 pages. ccs_tool(8) can be used to do some basic cluster.conf editing.
23
24 The cluster.rng schema is used to validate cluster.conf. Unrecognized
25 items will produce a warning during cluster startup, and invalid xml
26 structure will cause the cluster startup to fail. See ccs_config_vali‐
27 date(8) and ccs_config_dump(8).
28
29 A list of the valid items for your installed cluster version is held in
30 the file /usr/share/doc/cman-<version>/cluster_conf.html .
31
32
33
34 Cluster
35 The top level cluster section contains all other sections and has two
36 required attributes:
37
38 name The name of the cluster can be up to 15 characters long (16
39 including terminating null). It is important that this name be
40 unique among clusters on the same network.
41
42 config_version
43 The config_version specifies the revision level of the file and
44 should be increased each time the file is updated.
45
46 <cluster name="alpha" config_version="1">
47 </cluster>
48
49
50 Cluster Nodes
51 The set of nodes that make up the cluster are defined in the clustern‐
52 odes section which contains multiple clusternode sections. A clustern‐
53 ode has two required attributes:
54
55 name The node name should correspond to the hostname on the network
56 interface to be used for cluster communication.
57
58 nodeid The node id must be greater than zero and unique.
59
60 <cluster name="alpha" config_version="1">
61 <clusternodes>
62 <clusternode name="node-01" nodeid="1">
63 </clusternode>
64
65 <clusternode name="node-02" nodeid="2">
66 </clusternode>
67
68 <clusternode name="node-03" nodeid="3">
69 </clusternode>
70 </clusternodes>
71 </cluster>
72
73
74 Logging
75 Cluster daemons use a common logging section to configure their
76 loggging behavior.
77
78 <cluster name="alpha" config_version="1">
79 <logging/>
80 </cluster>
81
82 Global settings apply to all:
83
84 <logging debug="on"/>
85
86 Per-daemon logging_daemon subsections override the global settings.
87 Daemon names that can be configured include: corosync, qdiskd, groupd,
88 fenced, dlm_controld, gfs_controld, rgmanager.
89
90 <logging>
91 <logging_daemon name="qdiskd" debug="on"/>
92 <logging_daemon name="fenced" debug="on"/>
93 </logging>
94
95 Corosync daemon settings apply to all corosync subsystems by default,
96 but subsystems can also be configured individually. These include CLM,
97 CPG, MAIN, SERV, CMAN, TOTEM, QUORUM, CONFDB, CKPT, EVT.
98
99 <logging>
100 <logging_daemon name="corosync" subsys="QUORUM" debug="on"/>
101 <logging_daemon name="corosync" subsys="CONFDB" debug="on"/>
102 </logging>
103
104 The attributes available at global, daemon and subsystem levels are:
105
106
107 to_syslog
108 enable/disable messages to syslog (yes/no), default "yes"
109
110
111 to_logfile
112 enable/disable messages to log file (yes/no), default "yes"
113
114
115 syslog_facility
116 facility used for syslog messages, default "local4"
117
118
119 syslog_priority
120 messages at this level and up will be sent to syslog, default
121 "info"
122
123
124 logfile_priority
125 messages at this level and up will be written to log file,
126 default "info"
127
128
129 logfile
130 the log file name, default /var/log/cluster/<daemon>.log
131
132
133 debug="on"
134 a shortcut for logfile_priority="debug"
135
136
138 An explicit configuration for the default settings would be:
139
140 <logging to_syslog="yes" to_logfile="yes" syslog_facility="local4"
141 syslog_priority="info" logfile_priority="info">
142 <logging_daemon name="qdiskd"
143 logfile="/var/log/cluster/qdiskd.log"/>
144 <logging_daemon name="fenced"
145 logfile="/var/log/cluster/fenced.log"/>
146 <logging_daemon name="dlm_controld"
147 logfile="/var/log/cluster/dlm_controld.log"/>
148 <logging_daemon name="gfs_controld"
149 logfile="/var/log/cluster/gfs_controld.log"/>
150 <logging_daemon name="rgmanager"
151 logfile="/var/log/cluster/rgmanager.log"/>
152 <logging_daemon name="corosync"
153 logfile="/var/log/cluster/corosync.log"/>
154 </logging>
155
156 To include debug messages (and above) from all daemons in their default
157 log files, either of the following which are equivalent:
158
159 <logging debug="on"/>
160 <logging logfile_priority="debug"/>
161
162 To exclude all log messages from syslog:
163
164 <logging to_syslog="no"/>
165
166 To disable logging to all log files:
167
168 <logging to_file="no"/>
169
170 To include debug messages (and above) from all daemons in syslog:
171
172 <logging syslog_priority="debug"/>
173
174 To limit syslog messages to error (and above), keeping info (and above)
175 in log files (this logfile_priority setting is the default so could be
176 omitted):
177
178 <logging syslog_priority="error" logfile_priority="info"/>
179
181 /etc/cluster/cluster.conf
182 standard location of cluster configuration file
183
184 /usr/share/cluster/cluster.rng
185 standard location of cluster.conf schema
186
187
189 ccs_tool(8), ccs_config_dump(8), ccs_config_validate(8), cman_tool(8),
190 cman(5), qdisk(5), fenced(8), fence_node(8), dlm_controld(8), gfs_con‐
191 trold(8), rgmanager(8)
192
193
194
195
196cluster 2010-01-12 CLUSTER.CONF(5)