SUMMARY: Oracle problem after a restore
From: Benoît Audet (benoit.audet_at_carra.gouv.qc.ca)
Date: 04/25/05
- Previous message: Bernd Schemmer: "SCSI error codes"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Mon, 25 Apr 2005 14:30:45 -0400 To: <sunmanagers@sunmanagers.org>
First of all, I would like to thank all of those who responded to my
demand for help. Especially, Tommy Falsen David B Harrington, Bernd
Schemmer ans Scott Mickey, who pointed me to interesting directions.
Surprisingly, just before noon, I finally got from one of the DBAs that
they did a change last week to the server: they added 3 more Oracle
instances. The server asn't been rebootet before this time, of course,
and everything went well until yesterday.
We did a full restore, via NetBackup, on new disks, changing *nothing
else than the swap partition*, which is now twice the size it was before
the move.
Looking closely (two more "sunmanagers"'s mails about swap from me), I
finally saw that Oracle instances was taking much more swap than it was
before the restore. I just added another swap partition *just to see*
and TADAAAM! everything came up flawlessly!
Strange thing I suspect is 1) Oracle would need a serious fine-tuning
effort, and 2) it's strange that the server was running without problem
since the DBAs added 3 more instances, but they did add 3 more instances
on the server *without warn us* they did so!
Well, that's almost it. Thanks again to all who replied!
Ben Audet
My original post was:
Fellow managers,
I know that this is probably more an Oracle related problem and that
here isn't an Oracle mailing list, but since we don't know where to
look, I try here anyway.
We have a Sun Fire 280R with a StorEdge S1, runnung both Oracle 8.1.6
and 8.1.7.4. We had to perform a full restore on the server yesterday.
Despite the facts that we had a "good" full backup, the restore has been
applied successfully, and that the server came up very well after the
operation, Oracle don't work properly.
We have this message in the listener log file:
TNS-12500: TNS:listener failed to start a dedicated server process
TNS-12540: TNS:internal limit restriction exceeded
TNS-12560: TNS:protocol adapter error
TNS-00510: Internal limit restriction exceeded
Solaris Error: 12: Not enough space
Everywhere I look for an explanation, it points to the fact that the
server is "missing" memory and/or swap. But the settings are the same
as "before" we had to restore the server, except that the swap partition
is twice the size (we replaced internal disks with bigger ones, so we
had room for more swap).
The actual server have 1 Gb. of RAM, plus 4 Go. of swap:
oracledev# swap -l
swapfile dev swaplo blocks free
/dev/md/dsk/d2 85,2 16 8389632 8389632
oracledev# swap -s
total: 597896k bytes allocated + 4140144k reserved = 4738040k used,
209280k available
Here are the "/etc/system" values related to Oracle:
set shmsys:shminfo_shmmax=4294967295
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=100
set shmsys:shminfo_shmseg=10
set semsys:seminfo_semmap=64
set semsys:seminfo_semmns=4096
set semsys:seminfo_semmni=4096
set semsys:seminfo_semmnu=4096
set semsys:seminfo_semume=64
set semsys:seminfo_semmsl=250
set semsys:seminfo_semopm=100
set semsys:seminfo_semvmx=32767
set rlim_fd_max=2048
set rlim_fd_cur=4096
Well, this is almost it. Unfortunately, the Oracle knowledge we have
here is a little limited... ...so that's why I ask if some of you
managers would have a clue on what could help me to fix the problem.
Thanks in advance and, of course, will summarise.
Ben Audet
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers
- Previous message: Bernd Schemmer: "SCSI error codes"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|
|
