Charles,
Some months ago I had a similar problem. I wanted to
uninstall and reinstall the ASM instance. Unfortunately I don't remember the
specific steps I took but I believe I went through everything you did and then
tried to recreate it using the OUI and the installer kept finding pieces of
it as I would try. After some frustration I opened an SR with Oracle and
they provided the following steps, most of which it looks like you have
completed, nevertheless, here they are...
1) Shutdown all databases
that are running on the box using the +ASM instance for storage management
2)
Use emca -x <sid> to remove the Database Console for each instance
3)
Clear the existing raw disks
dd if=/dev/zero of=<raw device> bs=8192
count=1000
OR
drop the raw partition
4) Drop the old parameter file
(init<sid>.ora or spfile<sid>.ora)
Default location:
<ORACLE_HOME>\dbs
5) Delete / Remove any non ASM supporting
directories
BACKGROUND_DUMP_DEST
USER_DUMP_DEST
CORE_DUMP_DEST
etc
6)
Remove the ORATAB entry for the ASM instance
7) Remove the resource from
cluster ready services (CRS)
srvctl remove asm -n nodename
Ultimately that worked but I don't quite know which piece of it worked.
One thing I did do was significantly increase the byte count when running the dd
statement and that seemed to play a major part in getting things to work. The
other piece which I don't see mention of is unregistering the ASM Disk groups
from ASMLib. I don't know if you are using that and if that matters but I
believe I did that at one point also.
I'm
not sure this will help but I did go through it once before. Nevertheless it is
still a black box.
Thanks.
Bill Wagman
Univ. of California at Davis
IET Campus Data
Center
wjwagman@ucdavis.edu
(530) 754-6208
I posed this question over at the oracle ASM forum as well.
We
are trying to wipe out ASM. We have been trying to follow a couple metalink
notes, and threads both here and in oracle-l, not to mention reading some books
(K Gopal and Julian Dyke). Google is not giving us much help, either.
We
had a couple RAC databases, but those have already been removed, and the shared
ASM disks have been wiped clean (with dd). We have not touched the OCR or voting
disks directly at all, only via srvctl and the crs commands. I have removed as
many dependencies as I can think of (entry in oratab, running processes, ?dump
directories).
Oracle EE
10.2.0.3 running on
RHEL4. Below are a series of commands we have used. I attempted to find the "CRS
daemon log file", but it is not updated with any relevant information. The ocssd
log file has some interesting stuff, but nothing that leads me to a
solution.
<urbdb1> /u01/app/oracle: srvctl remove asm -n urbdb1
PRKS-1033 : Failed to remove configuration for ASM instance "+ASM1" on node "urbdb1" from cluster registry,
[PRKS-1023 : Failed to remove CRS resource for ASM instance "+ASM1" on node "urbdb1",
[CRS-0214: Could not unregister resource 'ora.urbdb1.ASM1.asm'.]]
[PRKS-1023 : Failed to remove CRS resource for ASM instance "+ASM1" on node "urbdb1",
[CRS-0214: Could not unregister resource 'ora.urbdb1.ASM1.asm'.]]
<urbdb1> /u01/app/oracle: crs_unregister ora.urbdb1.ASM1.asm
CRS-0214: Could not unregister resource 'ora.urbdb1.ASM1.asm'.
<urbdb1> /u01/app/oracle: oerr crs 214
214, 0, "Could not unregister resource '%s'."
// *Cause: There was an internal error while unregistering the resource.
// *Action: Check the CRS daemon log file.
<urbdb1> /u01/app/oracle: find . -type f -mmin -1 -exec ls -alt {} \;
-rw-r----- 1 oracle oinstall 183445 Jul 24 09:29 ./product/crs/cv/log/cvutrace.log.0
-rw-r--r-- 1 oracle oinstall 7072151 Jul 24 09:29 ./product/crs/log/urbdb1/cssd/ocssd.log
-rw-r--r-- 1 oracle oinstall 186 Jul 24 09:29 ./product/crs/log/urbdb1/client/css4348.log
-rw-r--r-- 1 oracle oinstall 809409 Jul 24 09:29 ./product/asm/network/log/listener.log
-rw-r----- 1 oracle oinstall 9292901 Jul 24 09:29 ./product/agent10g/network/log/sqlnet.log
As always, I am hoping the answer to this trickiness is easy and
staring me in the face. =)
--
Charles Schultz
