1OCF_HEARTBEAT_MPATHP(7)       OCF resource agents      OCF_HEARTBEAT_MPATHP(7)
2
3
4

NAME

6       ocf_heartbeat_mpathpersist - Manages SCSI persistent reservations on
7       multipath devices
8

SYNOPSIS

10       mpathpersist [start | stop | monitor | promote | demote | meta-data |
11                    validate-all]
12

DESCRIPTION

14       This resource agent manages SCSI persistent reservations on multipath
15       devices. "mpathpersist" from multipath-tools is used, please see its
16       documentation. Should be used as multistate (Master/Slave) resource
17       Slave registers its node id ("crm_node -i") as reservation key (
18       --param-sark ) on each device in the params "devs" list. Master reservs
19       all devices from params "devs" list with reservation "--prout-type"
20       value from "reservation_type" parameter. Please see man sg_persist(8)
21       and mpathpersist(8) for reservation_type details.
22

SUPPORTED PARAMETERS

24       binary
25           The name of the binary that manages the resource.
26
27           (optional, string, default "mpathpersist")
28
29       devs
30           Device list. Multiple devices can be listed with blank space as
31           separator. Shell wildcars are allowed.
32
33           (required, string, no default)
34
35       required_devs_no
36           Minimum number of "working" devices from device list 1) existing 2)
37           "mpathpersist --in --read-keys <device>" works (Return code 0)
38           resource actions "start","monitor","promote" and "validate-all"
39           return "OCF_ERR_INSTALLED" if the actual number of "working"
40           devices is less than "required_devs_no". resource actions "stop"
41           and "demote" tries to remove reservations and registration keys
42           from all working devices, but always return "OCF_SUCCESS"
43
44           (optional, string, default "1")
45
46       reservation_type
47           reservation type
48
49           (optional, string, default "1")
50
51       master_score_base
52           master_score_base value "master_score_base" value is used in
53           "master_score" calculation: master_score = master_score_base +
54           master_score_dev_factor * working_devs if set to bigger value in
55           mpathpersist resource configuration on some node, this node will be
56           "preferred" for master role.
57
58           (optional, string, default "0")
59
60       master_score_dev_factor
61           Working device factor in master_score calculation each "working"
62           device provides additional value to "master_score", so the node
63           that sees more devices will be preferred for the "Master"-role
64           Setting it to 0 will disable this behavior.
65
66           (optional, string, default "100")
67
68       master_score_delay
69           master/slave decreases/increases its master_score after delay of
70           "master_score_delay" seconds so if some device gets inaccessible,
71           the slave decreases its master_score first and the resource will no
72           be watched and after this device reappears again the master
73           increases its master_score first this can work only if the
74           master_score_delay is bigger then monitor interval on both master
75           and slave Setting it to 0 will disable this behavior.
76
77           (optional, string, default "30")
78

SUPPORTED ACTIONS

80       This resource agent supports the following actions (operations):
81
82       start
83           Starts the resource. Suggested minimum timeout: 30s.
84
85       promote
86           Promotes the resource to the Master role. Suggested minimum
87           timeout: 30s.
88
89       demote
90           Demotes the resource to the Slave role. Suggested minimum timeout:
91           30s.
92
93       notify
94           Suggested minimum timeout: 30s.
95
96       stop
97           Stops the resource. Suggested minimum timeout: 30s.
98
99       monitor (Slave role)
100           Performs a detailed status check. Suggested minimum timeout: 20s.
101           Suggested interval: 29s.
102
103       monitor (Master role)
104           Performs a detailed status check. Suggested minimum timeout: 20s.
105           Suggested interval: 60s.
106
107       meta-data
108           Retrieves resource agent metadata (internal use only). Suggested
109           minimum timeout: 5s.
110
111       validate-all
112           Performs a validation of the resource configuration. Suggested
113           minimum timeout: 30s.
114

EXAMPLE CRM SHELL

116       The following is an example configuration for a mpathpersist resource
117       using the crm(8) shell:
118
119           primitive p_mpathpersist ocf:heartbeat:mpathpersist \
120             params \
121               devs=string \
122             op monitor depth="0" timeout="20s" interval="29s" role="Slave" \
123             op monitor depth="0" timeout="20s" interval="60s" role="Master"
124
125           ms ms_mpathpersist p_mpathpersist \
126             meta notify="true" interleave="true"
127

EXAMPLE PCS

129       The following is an example configuration for a mpathpersist resource
130       using pcs(8)
131
132           pcs resource create p_mpathpersist ocf:heartbeat:mpathpersist \
133             devs=string \
134             op monitor depth="0" timeout="20s" interval="29s" role="Slave" \
135             op monitor depth="0" timeout="20s" interval="60s" role="Master" --master
136

SEE ALSO

138       http://clusterlabs.org/
139

AUTHOR

141       ClusterLabs contributors (see the resource agent source for information
142       about individual authors)
143
144
145
146resource-agents UNKNOWN           10/24/2018           OCF_HEARTBEAT_MPATHP(7)
Impressum