Personnel
- QC is Sandra
- DBA is Prabhu
- Sysadmin Joncarlo
Preparation for Swith Over
- (Prabhu) Copy all the cron jobs from Hammer to Steele.
- (Prabhu) Verify with sysadmin that patching of hammer is completedCopy all the scripts from Hammer to Steele and edit the necessary parameters.
- (Prabhu) Verify with Backup admin/Sysadmin that filesystem backup/snapshot is completed
- Verify with sysadmin about firewall rules, processes limits, Kernel parameters, and oracle limits are the same on hammer and steele.
During maintenance window
S = maintenance window + 30 mins 2 Hrs
- (S - 15 min m ProjManmgt/Prabhu) Set database marker.Make sure database marker is set prior to start of the patching.
- (S - m Prabhu) Verify with sysadmin that patching of hammer is completed
- (S - 10 min Programmer) Put all application servers into maintenance mode and shutdown Sakai on each one
- (S + 5m Prabhu) Backup the full database
- m Prabhu) Make sure all application servers are offline.
- (S + m Prabhu) Backup the full database on hammer.
- (S + m Prabhu) Make sure backup of full database on hammer is complete.
- (S + m Prabhu) Shutdown the standby database on Steele and Primary database on Hammer
- (S + m Prabhu) Inform Backupadmin/Sysadmin to start filesystem backup/snapshot on Hammer and Steele.
- (S + m Prabhu) Make sure filesystem snapshot is complete and Start the standby database on Steele and then Primary database on Hammer.
- (S + m Prabhu) Execute switchover procedure
- (S + m Sysadmin) Restart sakai on head node
- (S + m Prabhu/Sysadmin) Verify the application login
- (S + m Sysadmin) Restart sakai on remaining nodes
- (S + m Sysadmin) Take application servers out of maintenance mode
Fall-back Procedure (If needed) Time D is the decision time made to fallback
- (D + 0m m Sysadmin) Shutdown Sakai on smartsite-prod
- (D + m Prabhu ) Execute Switch back procedure
- (D + 20m Programmerm Sysadmin) Restart sakai on head node
- (D + 25m QCm Prabhu / Sysadmin) Verify the database marker is in placeapplication login
- (D + 40m Programmerm Sysadmin) Restart sakai on remaining nodes
- (D + 25m Programmerm Sysadmin) Take application servers out of maintenance mode