RE: Manual mem management in 10g

Before we set sga_target = 0, we were seeing "KGH: NO ACCESS" grow to over 50% of the shared_pool. Support would only say that this was used to temporarily transfer data from the buffer cache to the shared_pool, but they never did disclose why so much, or how, or when, or....

Would you really want most of your memory tied up in something called "NO ACCESS"? *grin*

-----Original Message-----

From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Allen, Brandon Sent: Friday, June 30, 2006 11:25 AM
To: topshot.rhit_at_gmail.com; oracle-l_at_freelists.org Subject: RE: Manual mem management in 10g

I'm growing concerned about the same thing after seeing some of the comments on this list. I just implemented my first 10g system (10.2.0.2 on AIX 5.3) about a month ago and I decided to go out on a limb and use the new gather_stats_job and sga_target. So far, our performance has been excellent and we haven't had any problems with memory errors, but after all the problems I've seen on this list I'm increasingly worried that I may be sitting on a time bomb and maybe I should change back to manual memory management before it explodes. I've just recently started keeping a closer eye on v$sgastat, hoping that I can catch any problems before they get out of hand. A couple odd things I've noticed - a lot of free memory in the shared pool, even though ADDM keeps telling me I need to increase my SGA, and a lot of memory allocated to "KGH: NO ACCESS": SQL> select * from v$sgastat where bytes > 10000000 order by 3;

POOL         NAME                            BYTES


------------ -------------------------- ----------

shared pool  private strands              11624448
shared pool  KQR M PO                     13381408
             log_buffer                   14700544
shared pool  obj stat memo                16219416
shared pool  ASH buffers                  16252928
java pool    free memory                  16777216
shared pool  Cursor Stats                 17322216
shared pool  kglsim heap                  17627904
shared pool  PCursor                      23200608
shared pool  CCursor                      29468216
shared pool  kglsim object batch          31205664
shared pool  library cache                31840456
shared pool  KGLS heap                    50571752
large pool   free memory                  62080224
shared pool  sql area                    176615248
shared pool  KGH: NO ACCESS              480967264
shared pool  free memory                 655483792
             buffer_cache               1241513984

Anyone else seeing the same pattern? Have an explanation?

Thanks,
Brandon

Privileged/Confidential Information may be contained in this message or attachments hereto. Please advise immediately if you or your employer do not consent to Internet email for messages of this kind. Opinions, conclusions and other information in this message that do not relate to the official business of this company shall be understood as neither given nor endorsed by it.

--

http://www.freelists.org/webpage/oracle-l

--

http://www.freelists.org/webpage/oracle-l Received on Fri Jun 30 2006 - 12:27:57 CDT