Document Type | Technical Information
Category | Backup/Recovery
Applicable Product Versions | T6FS07, T6FS07PS, T7FS02, 7FS02PS
Document Number | TBATI050
Overview
This document describes a method to perform failover of TSC without interruption in a TSC-OBSERVER configuration.
Method
Configuration Information
Server Configuration
Check OBSERVER CM Status
The OBSERVER status can be checked using the cmrctl show command.
- CM_NAME
This is the name set when registering the CM Resource on the Primary and Standby servers.
For example, CM01 refers to Node01, and CM02 refers to Node02. - PRI/TAR
Indicates the current startup status of the database.
PRIMARY means the database is running normally, and TARGET means it is running in Recovery mode on the Standby side.
Performing Failover
Force Shutdown of Primary and Verify Failover Operation
Force shutdown the Primary DB using the
tbdown abnormalcommand. This command forces shutdown without leaving logs.Then, check the OBSERVER status using the
cmrctl showcommand, and you will see that CM02, which was previously the Standby, has switched to Primary.Also, you can confirm that the former Standby DB automatically operates in NORMAL FAILOVER mode.
Verify Normal Startup of Standby Server
- When checking with cmrctl show, STATUS should be confirmed as Normal.
- When checking v$database, OPEN_MODE should be confirmed as READ WRITE.
Primary Server Recovery
Move Datafiles from Former Standby to Primary
Perform Hot Backup on the Standby datafiles and move them to the Primary server.
- Perform Begin Backup
- Move datafiles and Archive Logs to Primary (from Standby server to Primary server)
- Perform End Backup
Perform Recovery on Former Primary Server
- Start in Mount mode, perform standby controlfile and standby recover operations as in TSC configuration.
- Then shut down and start with tbboot recovery (start in Standby mode).
Check in OBSERVER
- You can confirm that CM01 (former Primary) is registered in TARGET status.