Switching services from active node to failovernode in cluster

Linux and Unix-Like Red Hat
1

Hi All,

I was asked to Switch the service from failover to active node as we need to reboot active node for regular maintenance. Unfortunately our senior admin left the organization. It would be helpful if some one answer this query. Detailed info with steps will be appreciated.

Here are the below details. I need to reboot lab1 and need to switch all services to lab2 or lab3.

I believe the /appfs is related to cluster. I'm bit rusty with cluster administration :o
Please help!!!:slight_smile:

#hostname
lab2

#clustat

 Member Name                              Status
    ------ ----                              ------
    lab1-int                             Online, rgmanager
    lab2-int                             Online, Local, rgmanager
    lab3-int                             Online, rgmanager
   
    Service Name         Owner (Last)                   State
    ------- ----         ----- ------                   -----
    app_gfs_svc         lab2-int                   started
    app_prod            lab1-int                   started
    app_test            (lab1-int)                 disabled
    app_upgrade         (none)                         disabled


  # df -TPH /appfs
  Filesystem    Type     Size   Used  Avail Use% Mounted on
  /dev/dm-50     gfs      32G    24G   8.3G  75% /appfs
2

This post was from a while ago, but i'll answer in case anyone finds this in a search looking for an answer.

The short answer is that there is nothing you need to do. If you shut one down the services should fail over automatically. The command to move a rgmanager service from one host to another is:

clusvcadm -r <service> -m <new_member>

For example:

clusvcadm -r app_prod -m lab3-int

All of this assumes that the services have been configured correctly and with best practice.

1 Like