1. Make sure all applications reading/writing from/to the databases on the PRIMARY server are disabled
2. Disable all the jobs on the PRIMARY server and wait for them to finish
3. Through Replication Monitor on the DISTRIBUTOR Stop Log Read, Stop Synchronization of 5 subscribers
4. Disable all the jobs on the DISTRIBUTOR server and wait for the jobs to finish
At this point PRIMARY and SUBSCRIBERS 1-5 should have identical data, with no new data and no new requests coming in.
The green light will be given to the IT team to copy over the data/log files to the drives on 3 PAR with identical drive names.
The IT team will take the PRIMARY cluster down (thereby stopping SQL Service and agent), copy over the files to the pre-created drives with allocated space. Once copying of data files is done, the cluster will be brought back up, and SQL
Service and Agent started for the PRIMARY Server.
Once databases are accessible, I will do the following:
1. Start Log Reader Agent
2. Start synchronizing SUBSCRIBERS 1-5
3. Enable jobs on DISTRIBUTOR
4. If replication is working, jobs on PRIMARY will be enabled, and new data will be allowed to come in.
My questions are as follows:
1. Is there someting I am missing in the plan? Should I be doing something else?
2. Am I doing any extra/unnecessary steps?
3. Will the DISTRIBUTOR complain if the PUBLISHER goes down? If so, should I just take it offline during this process?
4. Is there any chance that Replication can break during/after this process?
In our case, however, SQL Service was stopped. This allowed us to easily copy over the database files from one drive to another (with the same name). Once copying was over,SQL Service was started, and everything was working fine.