This article provides some scenarios for automation and recovery of DB2 resources.
Task | DB2 automation and recovery scenarios |
---|---|
Startup | DB2 is started after RRS but before LDAP, NFS, and WebSphere Application Server for z/OS. |
Shutdown | Reverse of startup sequence. |
Handling in-flight or indoubt transactions if there is a failure | Use the RRS panels
to resolve. See z/OS MVS Programming: Resource Recovery. The
RRS panels are the preferred way to resolve DB2 indoubts because they allow
you to view all resource managers that have an interest in the transaction.
However, you can also use DB2 to resolve indoubts. You can issue the command:DISPLAY THREAD(*) TYPE(INDOUBT)to display DB2 information about the indoubt threads it knows about (if there are too many, you can go into S.LOG to view the information). This display will give you a DB2 identifier called a "nid". Copy the nid and paste it into this command: -RECOVER INDOUBT (RRSAF) ACTION(COMMIT) NID(B1D379D17ED6CF900000009401010000)where the nid is the one that you cut from the display command. You can issue this command to roll back the transaction: -RECOVER INDOUBT (RRSAF) ACTION(ABORT) NID(B1D379D17ED6CF900000009401010000) |
How to determine if DB2 is running | Use the display command to display the DB2 address space. |
What happens to WebSphere Application Server for z/OS if DB2 goes down? | WebSphere Application Server for z/OS continues to run. WebSphere Application Server for z/OS does not require restarting in this scenario. |
What happens to other subsystems if DB2 goes down? | Not applicable |
Where to find more information | See the DB2 books at the following Internet location:http://www.ibm.com/servers/eserver/zseries/zos/ |