wd_cli
Name
wd_cli -- Watchdog Command Line Interface (CLI) utility for the external health-check system integration.Description
wd_cli Provides a convenient and easy to use interface to interact with watchdog and perform health check related node operations.
Operations
-
-i
--info -
Get the node status for node(s) based on node search criteria
-
-I
NEW_STATUS
--inform= NEW_STATUS -
Update the watchdog node status. Allowed values are DEAD and ALIVE
Node Search Options
-
-a
--all -
Select all nodes (only available with --info operation)
-
-n
WATCHDOG_NODE_ID
--node-id= WATCHDOG_NODE_ID -
Select watchdog node by node_id.
-
-N
NODE_NAME
--node-name= NODE_NAME -
Select watchdog node by name.
-
-H
HOSTNAME
--node-host= HOSTNAME -
Select watchdog node by hostname.
-
-P
PORT_NO
--node-port= PORT_NO -
Select watchdog node by port number.
Options
-
-k
KEY
--auth-key= KEY -
Watchdog auth key. Specifying auth-key over rides the pgpool.conf->wd_authkey value.
-
-f
CONFIG_FILE
--config-file= CONFIG_FILE -
Specifies the pgpool.conf file.
-
-s
SOCKET_DIR
--socket-dir= SOCKET_DIR -
Path to the watchdog IPC socket directory. Specifying socket-dir over rides the pgpool.conf->wd_ipc_socket_dir value.
-
-p
PORT_NO
--ipc-port= PORT_NO -
Port number of watchdog IPC socket. Specifying ipc-port over rides the pgpool.conf->wd_port value.
-
-m
message_string
--message= message_string -
Optional message string to be passed to Pgpool-II along with the node status.
-
-v
--verbose -
Enable verbose messages.
-
-V
--version -
Prints version information.
-
-d
--debug -
Enable debug output.
-
-h
--help -
Prints the help for wd_cli .
Example
When some external health-check system is configured to monitor the health of Pgpool-II cluster and it detects a node failure. wd_cli utility can be used to pass this information to Pgpool-II cluster so that it can take the appropriate actions.
#inform Pgpool-II about node named RN-110 failure
wd_cli --inform DEAD -N 'RN-110' -p 9001 -m 'Pgpool node with name RN-110 is not reachable'
#inform Pgpool-II about node-id 10 failure
wd_cli --inform DEAD -n 10 -m 'Pgpool node with ID 10 is not responding'
Similarly to inform Pgpool-II about the node is alive again.
#inform Pgpool-II about node-id 10 re-joining
wd_cli --inform ALIVE -n 10 -m 'Pgpool node with ID 10 is reachable again'
wd_cli can also be used to get the current status of cluster nodes.
wd_cli -f simple_conf/watchdog/pgpool_wd1.conf --info -v --all Total Watchdog nodes configured for lifecheck: 2 ***************** Node ID: 0 Node Status code 4 Node Status: LEADER Node Name: localhost:9991 Linux localhost.localdomain Node Host: localhost Node WD Port: 9001 Node Pgpool Port: 9991 Node ID: 1 Node Status code 7 Node Status: STANDBY Node Name: localhost:9992 Linux localhost.localdomain Node Host: localhost Node WD Port: 9002 Node Pgpool Port: 9992