Re: TNS-12500: TNS:listener failed to start a dedicated server process

From: joel garry <joel-garry_at_home.com>
Date: Thu, 8 Jan 2009 14:17:59 -0800 (PST)
Message-ID: <4670563e-1a7e-4887-805c-0cdc8b21beba_at_35g2000pry.googlegroups.com>



On Jan 8, 12:25 pm, shweta.kapar..._at_googlemail.com wrote:
> On Jan 8, 8:22 pm, Steve Howard <stevedhow..._at_gmail.com> wrote:
>
>
>
>
>
> > On Jan 8, 2:26 pm, shweta.kapar..._at_googlemail.com wrote:
>
> > > On Jan 8, 2:34 pm, Steve Howard <stevedhow..._at_gmail.com> wrote:
>
> > > > On Jan 8, 9:29 am, shweta.kapar..._at_googlemail.com wrote:
>
> > > > > On Jan 8, 1:59 pm, Steve Howard <stevedhow..._at_gmail.com> wrote:
>
> > > > > > On Jan 8, 8:57 am, shweta.kapar..._at_googlemail.com wrote:
>
> > > > > > > Hi All
>
> > > > > > > We are badly being hit by the following error:
>
> > > > > > > 08-JAN-2009 12:43:05 * (connect_data=(server=dedicated)
> > > > > > > (service_name=XX)(INSTANCE_ROLE=primary)(failover_mode=(type=session)
> > > > > > > (method=basic)(backup=YY)(retries=180)(delay=5))(CID=(PROGRAM=oracle)
> > > > > > > (HOST=ZZ)(USER=oracle))(INSTANCE_NAME=AAAA)) * (ADDRESS=(PROTOCOL=tcp)
> > > > > > > (HOST=<IP>)(PORT=<PORT>)) * establish * BB * 12500
> > > > > > > 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
> > > > > > >     HPUX Error: 11: Resource temporarily unavailable
>
> > > > > > > This is repeating at the following interval. ( as a sample.).. It is
> > > > > > > intermittent.
>
> > > > > > > Start                    End
> > > > > > > 08/01/2009 12:40:03     08/01/2009 12:40:13
> > > > > > > 08/01/2009 12:41:03     08/01/2009 12:41:04
> > > > > > > 08/01/2009 12:42:04     08/01/2009 12:42:04
> > > > > > > 08/01/2009 12:43:04     08/01/2009 12:43:04
> > > > > > > 08/01/2009 12:43:05     08/01/2009 12:43:05
>
> > > > > > > In between above intervals connections are successful
>
> > > > > > > We are using "dedicated Server" connection.
>
> > > > > > > CPU /Mem utlisation is normal. No change seen in the pattern.
>
> > > > > > > Regards
>
> > > > > > > Shweta.
>
> > > > > > Hi Shweta,
>
> > > > > > How do your OS kernel parameter limits look, i.e. open files, max
> > > > > > processes, etc.?
>
> > > > > > Regards,
>
> > > > > > Steve- Hide quoted text -
>
> > > > > > - Show quoted text -
>
> > > > > Hi Steve
>
> > > > > We have 05 databases running in Host -A
>
> > > > > I have taken below o/p  from 1st database. We ra facing this issue in
> > > > > all the databases almost at the same time.
>
> > > > > SQL>select * from v$resource_limit;
>
> > > > > RESOURCE_NAME                  CURRENT_UTILIZATION MAX_UTILIZATION
> > > > > INITIAL_AL  LIMIT_VALU
> > > > > ------------------------------ ------------------- ---------------
> > > > > ---------- ----------
> > > > > processes                                      135
> > > > > 256       1000       1000
> > > > > sessions                                       134
> > > > > 252       1105       1105
> > > > > enqueue_locks                                   72
> > > > > 361      14154      14154
> > > > > enqueue_resources                               72
> > > > > 72      50000  UNLIMITED
> > > > > ges_procs                                      133
> > > > > 253       1001       1001
> > > > > ges_ress                                     37791
> > > > > 50079      90367  UNLIMITED
> > > > > ges_locks                                    28140
> > > > > 42011      96306  UNLIMITED
> > > > > ges_cache_ress                                5361
> > > > > 8522          0  UNLIMITED
> > > > > ges_reg_msgs                                   222
> > > > > 250       2230  UNLIMITED
> > > > > ges_big_msgs                                    17
> > > > > 28       2230  UNLIMITED
> > > > > ges_rsv_msgs                                     0
> > > > > 0       1000       1000
> > > > > gcs_resources                                72634           81356
> > > > > 122432     122432
> > > > > gcs_shadows                                  26643
> > > > > 26663      67337      67337
> > > > > dml_locks                                        0
> > > > > 36       1024  UNLIMITED
> > > > > temporary_table_locks                            0               2
> > > > > UNLIMITED  UNLIMITED
> > > > > transactions                                    14
> > > > > 66       1215  UNLIMITED
> > > > > branches                                         2
> > > > > 28       1215  UNLIMITED
> > > > > cmtcallbk                                        4
> > > > > 29       1215  UNLIMITED
> > > > > sort_segment_locks                             134             137
> > > > > UNLIMITED  UNLIMITED
> > > > > max_rollback_segments                           11
> > > > > 11        244        244
> > > > > max_shared_servers                               1
> > > > > 1         20         20
> > > > > parallel_max_servers                             4
> > > > > 121        241        241
>
> > > > > 22 rows selected.
>
> > > > > Thanks.
>
> > > > Hi Shweta,
>
> > > > I meant on the OS itself, i.e. HP-UX kernel limits.  I have seen
> > > > exceptions such as the one you posted when processes, open files, etc.
> > > > were approaching/at their maximum limit as allowed/configured by the
> > > > system administrator.
>
> > > > HTH,
>
> > > > Steve- Hide quoted text -
>
> > > > - Show quoted text -
>
> > > Thanks Mark and Steve,
>
> > > Metalink Document 146859.1 describes the exact problem
>
> > > and suggested fix is -
>
> > > "The problem can be avoided by raising the value of UNIX kernel
> > > parameter:
>
> > >         maxuprc - max number of processes per user
>
> > > Refer to the HP Operating system documentation for detailed steps."
>
> > > I need some clarification on this :
>
> > > a) Here "max number of processes per user"  : does it referring to
> > > unix   "oracle" user.
> > > because each processes are owned /fork by oracle user at operating
> > > system level?
>
> > > b) Any issues/side effects if we change the maxuprc.
>
> > > Regards
>
> > As long as the server can handle that many processes, it is just a
> > matter of increasing them.  AFAIK, this can be done live, so ask your
> > HP-UX system administrator to bump them up.
>
> > Alos, maybe you have idle connections that could be released (a good
> > admin is probably going to ask you that before he/she bumps up the
> > number)?- Hide quoted text -
>
> > - Show quoted text -
>
> Thanks a lot Steve.

While it is usually either simply number of processes or things keep making too many processes, the fact that all your instances do this at once could be something a little more obscure, like running out of swap space or tcp issues. Watch it carefully! Easy to goof with many instances, especially when things go to heck because these erroring processes aren't cleaning up correctly.

How are you accessing the db - jdbc, odbc, sqlplus, what?

Also post your exact os, platform, oracle versions.

jg

--
_at_home.com is bogus.
http://blog.wired.com/27bstroke6/2009/01/professed-twitt.html
Received on Thu Jan 08 2009 - 16:17:59 CST

Original text of this message