ora-01102 How can oracle db behave like this?

We have two 8.1.6 dataases NJ01 (33 GIG) and NJ02 (7GIG) on Win2000 server. Yesterday there was some data corruption on disk (3 index datafiles gone for NJ01 and 1 rollback datafile gone for NJ02). I did offline drop the datafiles and opened the databases. Today i was trying to fix the problem by dropping/recreating affected table spaces.

1. To fix NJ01, I set oracle_sid=NJ01 and get a list of files on NJ01. Then I gave shutdown, killed the service and began a cold backup of OS files.

2. For NJ02, metalink recommened to fix the rollback tablespace while the database is up. So I connected (internal/pwd@nj02) without realizing that oracle sid is set to NJ01. But note that NJ01 was shutdown.

3. This is where strange things happened. NJ02 opened without incident
with the follow results.

select name from v$database;

NJ01

select instance_name from v$instance;

NJ02

Select name from v$datafile;

(all the NJ01 datafiles)

Did oracle open NJ02 with NJ01 parameter file? I was shocked and I shutdown the database by stopping the service.

By now i thought I hosed both databases. I discontinued cold backup of NJ01 and brought up the service. Database did not open. I tried through svrmgrl. One datafile complained .. could not lock the file.. so i did a offline drop on that file and opened NJ01.

The I started the service for NJ02. Database did not open. Then i tried to open it through svrmgrl... i got the error..

ora-01102 Can not open the database in exclusive mode..

GURUS.. PLEASE TELL ME WHAT HAPPNED HERE AND WHETHER I CAN DO ANYTHING TO OPEN NJ02.. THANKS..

Best is:

c:\set oracle_sid=NJ01
c:\svrmgrl
svmgrl>connect internal
svmgrl>shutdown immediate / shutdown abort
svmgrl>exit

c:\set oracle_sid=NJ02
c:\svrmgrl
svmgrl>connect internal
svmgrl>shutdown immediate / shutdown abort
svmgrl>exit

Once you do this then:

c:\set oracle_sid=NJ01
c:\svrmgrl
svmgrl>connect internal
svmgrl>startup mount
svmgrl>alter database open;
svrmgr>exit

c:\set oracle_sid=NJ02
c:\svrmgrl
svmgrl>connect internal
svmgrl>startup mount
svmgrl>alter database open;
svmgrl>exit

If you have any problems during these procedure you can post here, as in which step you encountered this error and error code and description. Meantime you can also check your log and trace file.