Hi,

I am currently running two Oracle 10G instances from a single Solaris
M4000 server. Every few days one of the instances crashes with the
following error and has to be restarted.

ORA-00470: LGWR process terminated with error
PMON: terminating instance due to error 470

I have spend weeks running various trace files etc with Oracle support
and they are basically clueless as to what is the cause. They are
saying it is a Solaris OS issue but surely this would affect both
instances and not just one.

Essentially something is killing the LGWR process and the instance is
shutting itself down. I think the way ahead is to try and find out
what is killing this process but I'm not sure how to go about this and
worried that any logging may degrade server performance.

Can anybody offer any suggestions ?

Thanks in advance
John

On Wed, 10 Mar 2010 04:07:48 -0800, Johne_uk wrote:

Let me see whether I understood you correctly: Oracle support is trying
to fix your issue for weeks, they presumably have all the trace files,
core files and RDA output but are unable to determine the exact cause of
the problem despite having all this information and you are asking us to
guess what the problem is without even knowing the version of the RDBMS
and Solaris? No problem, I can do that. The answer to your problem is the
number 42.


--
http://mgogala.freehostia.com

On 10 Mar, 13:48, Mladen Gogala <gogala.mla... (AT) gmail (DOT) com> wrote:
Actually I was looking for some assistance on determining what is
killing the LGWR process. Oracle do not provide OS support ! I thought
the question was clear enough but obviously not. Thanks for the
comment though as its must have wasted five minutes.

On Mar 10, 9:02*am, Johne_uk <edg... (AT) tiscali (DOT) co.uk> wrote:
I understand you are looking for assistance however you failed to
provide sufficient information in your vague problem description to
allow anyone to help. Yes, the LGWR process is dying; is it
accompanied by an ORA-07445 error? Is there any indication in the
resulting trace files that Solaris is at fault? Which release of
Oracle are you running? Yes, I know, you blurted out the wonderful
marketing speak of 10g but that umbrella covers two major releases
plus 8 patch levels with major changes between the two so please do us
a favor and take the time and effort to post that information to 4 or
5 numbers (10.1.0.4, 10.2.0.3, etc.). Also post the version of
Solaris this database is running on, including the curent patch level
(a uname -a can provide that information). As noted in a previous
post you've supplied far more information to Oracle and they,
apparently, can't solve the problem; why would you even think that
giving us such sketchy details would allow us to magically provide an
answer that Oracle (who has far more data than we do) cannot?

This is a help forum, not a carnival sideshow; we cannot pull answers
out of nothing more than thin air and thin air, really, is all you've
provided us. You have no right to complain regarding the response
first given you since you ambiguously described your problem and
expected detailed assistance to be forthcoming.

Supply much more information and you may find better help with your
issue.


David Fitzjarrell

On 10 Mar, 14:45, ddf <orat... (AT) msn (DOT) com> wrote:
Understood David I will follow up with more information
cheers

On 10 Mar, 15:02, Johne_uk <edg... (AT) tiscali (DOT) co.uk> wrote:
It's very strange behaviour from Oracle support, after send a trace,
log file and more and more, they don't tell anything...for me are very
busy with Sun's fusion or with your trimarano BMW Oracle in New
Zeland.
Version solaris?

Here's "something" to start with:

http://www.orafaq.com/forum/t/23468/2/

Please check that the DB is patched and updated with all the latest
patches (one-off as well as cpu patch).

Best of luck in your debugging efforts.

Regards

Eugene Pokopac (Oracle DBA - Tucker, GA)

On Mar 10, 9:02*am, Johne_uk <edg... (AT) tiscali (DOT) co.uk> wrote:
If you have access to Oracle support (as you stated), you should be
able to review the following Metalink documents:
Doc ID 1022782.6 - ORA-447, ORA-470 Found in PMON Trace File, Database
Crashed
Doc ID 431246.1 - Pmon terminated instance due to LGWR termination
ORA-470
Doc ID 874832.1 - LNS: Standby redo logfile ,DB Crash
Doc ID 745198.1 - RAC DATABASE HANGS WHEN RMAN IS STARTED
Doc ID 1029808.6 - BACKGROUND PROCESS DIES, NO TRACE FILE GENERATED
Doc ID 564830.1 - ORA-00470 Error Creating 20GB Tablespace
Doc ID 1016959.102 - HP-UX: DB CRASH. ALERT LOG SHOWS NO ERRORS, BUT
TRC FILES CREATED.

Some of the documents are specific to other operating systems, but may
provide you with clues that will help resolve the problem. One of the
articles mentioned that the problem could be caused by a slow file
system. Several of the articles indicated that the error may be
caused by exceeding the operating system limit for the maximum number
of open files.

Charles Hooper
Co-author of "Expert Oracle Practices: Oracle Database Administration
from the Oak Table"
http://hoopercharles.wordpress.com/
IT Manager/Oracle DBA
K&M Machine-Fabricating, Inc.

On Mar 10, 4:07*am, Johne_uk <edg... (AT) tiscali (DOT) co.uk> wrote:
Charles gave more details, but let me emphasize this: if the issue is
one of OS resources, it will surely affect one more than the other,
possibly the second one started.

You might want to google for dtrace, especially as it applies to
Oracle. Don't have time to look just now, but IIRC Tanel Poder, James
Morle and Kerry Osborne have some things about dtrace on their blogs.

Hey, oracle wants sun admins for aces: http://www.sun.com/bigadmin/home/index.jsp

Performance of a crashed instance is always the worst degradation.

jg
--
@home.com is bogus.
http://stevejmitchell.blogspot.com/2...astic-art.html

On Wed, 10 Mar 2010 09:02:33 -0800, joel garry wrote:

I beg to differ. If the database crashes, all your queries finish
instantly. I would even suggest that database crash is the ultimate thing
in application tuning.



--
http://mgogala.byethost5.com