To ensure successful stateful failover system setup, do not interrupt execution of the ms_recovery.sh script.
Running the script might take a long period of time, because the local disk partitions must be initialized and synchronized during the setup process. The time it takes to initialize and synchronize the local disk partitions depends on the disk size and network speed. For example, it takes about 20 minutes to initialize 100 GB of disk space over a Gigabit network, and using a 10 Gigabit network can speed up initialization.
Table-1 lists the parameters in the ms_recovery.sh script.
Table-1 Parameters in the ms_ recovery.sh script
|
Number |
Parameter |
Description |
|
$1 |
peer-node-ip |
Enter the management network IP address of the peer CVM host (replacement host). |
|
$2 |
peer-node-root-password |
Enter the root user password of the peer CVM host. |
|
$3 |
controller-root-password |
Enter the root user password of the quorum host. Do not set this parameter if you use ping quorum. |
Log in to the primary CVM host as the root user.
Run the ms_recovery.sh script, and then enter yes.
root@cvknode1:~# ms_recovery.sh 172.16.60.51 123456 123456
Please confirm to recovery 172.16.60.51 and format peer's /vms partition [yes/no]:
yes
------------------------Ip address Check------------------
Verify the script execution result.
If the CVM Master/Slave Restore Success message is displayed, the CVM stateful failover system has been recovered.
If error messages are displayed, the host has aborted the execution of the script because of errors. You can check the /var/log/cvm_master.log and /var/log/cvm_master_slave.log files for detailed error information.
Log in to CVM.