Hi,
Following is our current configuration:
- Server A , Failover-cluster, SQL Server 2008 sp3, Databases A (All databases are in read-write mode)
- Server B , Failover-cluster, SQL Server 2008 sp3, Databases A1, B, C (All databases are in read-write mode)
- Server C , Standalone, SQL Server 2008 sp3, Databases A1, B (A1 is in standby read-only mode, B in read-write mode))
- Server D , Standalone, SQL Server 2008 sp3, Databases A1, B, C (A1, B databases in restoring mode, C is in read-write mode)
- Server E , Standalone, SQL Server 2008 sp3, Databases A (A is in restoring mode)
Setup/Processes
- Transactional replication from Server A (DB A) to Server B(DB A1)
- Log shipping from Server B(DB A1) to Server C(DB A1)
- Transactional replication from Server B (DB B) to Server C(DB B)
- Log shipping from Server B(DB A1, B, C) to Server D(DB A1, B, C) (This is remote failover machine for disaster recovery)
- Log shipping from Server A(DB A) to Server E(DB A)
D is used for read-only reports for few days of month in remote location and E is remote failover machine. A, B, C are real production machines
We are planning to upgrade to SQL Server 2012 using following steps:
- Do pre-install steps (Run upgrade advisor, backup, checkdb etc.)
- Upgrade Passive cluster of Server A
- Failover to other node and upgrade 2<sup>nd</sup> node of cluster of Server A
- Upgrade databases on Server E
- Upgrade Passive cluster of Server B
- Failover to other node and upgrade 2<sup>nd</sup> node of cluster of Server B
- Upgrade databases on Server C
- Upgrade databases on Server D
I will be testing the plan on virtual machines in next few days but was wondering if someone can give an recommendation on what would be the best way to do this. Will log shipping, replication etc. will be affected and require setting them again or we can avoid resetting and able to upgrade, any particular issue to watch for cluster etc.