Procedure for the failover , when the source IP is changing (4331387)

Basically we have two nodes. One source and one target, from a Shareplex point of view. We need to be able to do a fail/switchover on the source node (which is our core business system), preferably without Shareplex going into a spin.

The source DB's file systems are mirrored, what happens in case of failover is that we bring all the applications down, attach the
mirrored file systems to the backup node and start everything up. All interfaces and applications use a DNS entry (billing), what we do is
change the IP address in the DNS from the primary node to the backup node, so that everybody connects to the live system. My understanding is that Shareplex is not able to handle the IP address change in the DNS and that is why we need this special procedure.
--source node:
DNS entry: Name: billing.dutchtone.nl Address: XXXXXXXX
Primary node: Name: bill-sp.dutchtone.nl Address: XXXXXXXX
Backup node: Name: bill-be.dutchtone.nl Address: XXXXXXXX
OS: SunOS bill-sp 5.9 Generic_118558-26 sun4u sparc SUNW,Sun-Fire-15000
INSTANCE_NAME HOST_NAME VERSION
-------------------------------------------------------------------------------------------------
BSCS bill-sp 9.2.0.6.0

SPLEX: bill-sp:splex:/prod45 [BSCS] $ env|grep DIR
SP_SYS_VARDIR=/sp02/var
SP_SYS_BINDIR=/sp01/prod45/bin

--target node:
Name: eai4rep.dutchtone.nl Address: XXXXXXXXX
SunOS eai4rep 5.8 Generic_117350-28 sun4us sparc FJSV,GPUZC-M

INSTANCE_NAME HOST_NAME VERSION
-------------------------------------------------------------------------------------------------
BSCSREP1 eai4rep 9.2.0.6.0

SP_SYS_VARDIR=/sp02/var45
SP_SYS_BINDIR=/sp01/prod45/bin

I already mentioned that the source system is our core business system,so what we want to achieve through failover should be obvious.

Setting up a virtual IP is not a choice here and source IP will be changing every time during the fail over.

Procedure for the failover , when the source IP is changing :

ON SOURCE :
1. Quiet the database on Source (Disable user access).
2. Issue flush on source:
sp_ctrl> flush o.<source sid>
3. Wait till number of messages zero in capture and export queues.
4. Make sure that Post is stopped due to flush on the TARGET, this will indicate that there is no data in the queues.
5. Shutdown Shareplex on SOURCE and TARGET.
6. cd $PRODDIR/bin
./qview
qview> qsetup <port#>
qview> list --this will show all queues on the source box
qview> deleteq C
qview> deleteq X - delete all export queues one by one if you have multiple export queues

ON TARGET:
7. Shareplex must be shutdown cd $PRODDIR/bin
./qview
qview> qsetup <port#>
qview> list
qview> deleteq P (delete all post queues by repeating the command)
qview> exit
8. Login SQL*Plus as shareplex user to truncate table Shareplex_Trans:
./sqlplus splex_user/splex_pwd
SQL> truncate table SHAREPLEX_TRANS;
SQL> exit
9. Perform the failover
10. Make sure that new IP/hostname can be resolved on the SOURCE and on the TARGET.

11. Make sure that your UNIX environment SP_SYS_HOST_NAME on SOURCE and TARGET to reflect the new IP/Hostname.

12. Startup Shareplex on the Source and Target.

13 Activate the config on the SOURCE, this will build the new queues with new routing info. Shareplex will lock the objects briefly for route changes.

14. Enable database user access from the new SOURCE.

Note: This is the same procedure to fail back, as the IP address in the only change from existing source during any failover.