Re: PGA Memory Leak - crashing Server

  • From: "LS Cheng" <exriscer@xxxxxxxxx>
  • To: bnsarma@xxxxxxxxx
  • Date: Wed, 19 Sep 2007 23:50:34 +0200

Hi

Do you know what is the simple change? Also do you see any ORA- error?

Thanks

--
LSC



On 9/19/07, BN <bnsarma@xxxxxxxxx> wrote:
>
> Greetings
>
> HP-UX 11.11, 4X4 Server
>
>
>
> Oracle version:
>
> Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - 64bi
> PL/SQL Release 10.2.0.1.0 - Production
> CORE    10.2.0.1.0      Production
> TNS for HPUX: Version 10.2.0.1.0 - Production
> NLSRTL Version 10.2.0.1.0 - Production
>
>
> PARAMETER                 VALUE
> ------------------------- ---------------
> _pga_large_extent_size    1048576
> _pga_max_size             209715200
>
> pga_aggregate_target                 big integer 100M
>
>
> We do see warnigns in AWR reports  to bump pga. we haven't done that.
> Could smaller pga cause this memory leak?
>
>
> We started with this version, plans are there to go to the Latest Patch
> Set. But for now we are stuck with this
>
> DEV folks did a release last Friday, its a simple change in  a package
>
> Problem:
>
> One process at a time goes upto  715 MB
> anothher process starts goes up to 715 MB
> another process start goes upto 715 MB
> . . . . . .
>
> Until system Crashes
>
> We Backeduout the code, we are back to normal.
>
> We have almost simialr load in one of our DEV server (4X4), We tried same
> code, we are not able to reproduce the issue.
>
> Only other option for us is to disable PGA and go for manual memory
> allocation by define sort_* values
>
> We have an SR open, we gave all the dumps oralce asked before we backed
> out the code in prod.
> Now oracle wants HEAP dumps, but managment doesn't want this  faulty code
> in prod until we fix the issue.
>
> The same code works fine in DEV. Server.
>
> Any Ideas and suggestions are appreciated
>
> Regards
> BN
>
>
>
> --
> Regards & Thanks
> BN
>

Other related posts: