This instruction presumes the availability of off-server/site backups of the following directories so that the files of those directories could be copied back to the server being recovered:
Recover Backup Server
Prepare the following:
Install a clean installation of CentOS 64bit
Configure networking to match the previous configuration
Ensure you can ping compute resources and Control Panel over the management and backup networks
Update CentOS components to the latest yum update.
Download and install the OnApp YUM repository file:
rpm -Uvh http://rpm.repo.onapp.com/repo/onapp-repo.noarch.rpmCODE
Install OnApp Backup Server installer package. Please, reboot if requested by the installer due to SELINUX, and then continue:
#> yum install -y onapp-bk-installCODE
Run the Backup Server installer to install the latest OnApp release:
If required, indicate the build number of the release version you want to install:
/onapp/onapp-bk-install/onapp-bk-install.sh -v <ONAPP_VERSION>CODE
Run the Backup Server configuration script to register it with the controller:
#>/onapp/onapp-bk-install/onapp-bk-config.sh -h <CP_HOST_IP> -p [BK_HOST_IP] -f <FILE_TRANSFER_SERVER_IP>CODE
- CP_HOST_IP- is the IP address of the Control Panel server
BK_HOST_IP - is the IP address of the Backup Server
FILE_TRANSFER_SERVER_IP - is the IP address of the server that will hold your backups and templates
To ensure the SSH keys are correct, import the following from the backup:
/root/.ssh/authorized_keys /root/.ssh/id_rsa /root/.ssh/id_rsa.pubCODE
Make sure you can access the backup server via SSH from the controller and the hypervisors. You should be able to connect through SSH as an OnApp user if the keys are working correctly.
Restore backup and template files to the new installation:
Basing on server configurations it’s possible that a separate block device is being used for backups. If so, this block device needs to be mounted under /onapp and all files transferred over to the block. If the /onapp/backups directory does not exist it can be created at this time.
Restore primary storage targets on the server.
The primary storage targets need to be mounted on the backup server.
If the environment is using iSCSI this can be done using the following commands:
Restore the /etc/iscsi/initiatorname.iscsi #> iscsiadmin -m discovery -t sendtargets -p [SAN IP] #> iscsiadmin -m node -lCODE
If the environment has FC/FCoE directly attached, you will need to ensure the block devices are visible and multipath is enabled:
#> service multipathd start #> multipath -llCODE
Once this is complete you can run the pvs command. You should see the primary storage targets, their identifiers will be in the following format: onapp- xxxxxxxxxxx.
- Ensure root access is available via SSH:
If the PermitRootLogin in the /etc/ssh/sshd_config is set to no, change it to yes to allow access to the backup server from the controller.
Restart services for changes to take effect:
service onapp stop service onapp startCODE