RE: slapd perf

["Howard Chu" <hyc@highlandsun.com>]

The relevant numbers are whatever is in use by
tests/scripts/test008-concurrency in the OpenLDAP source. The actual numbers
are meaningless
for the message I posted. All that's important is the relative performance,
since the test data sets are the same in both cases. The new code with
the optimizations I've committed is over 2X faster than the old code.

  -- Howard Chu
  Chief Architect, Symas Corp.       Director, Highland Sun
  http://www.symas.com               http://highlandsun.com/hyc
  Symas: Premier OpenSource Development and Support

> -----Original Message-----
> From: anish sheth [mailto:asheth@cisco.com]
> Sent: Monday, December 31, 2001 1:57 PM
> To: Howard Chu; openldap-devel@OpenLDAP.org
> Subject: Re: slapd perf
>
>
> hi,
>
> can you please provide more data as to how many entries in the directory
> etc.. for this search numbers?
>
> Thanks,
> -anish
>
> At 01:43 PM 12/31/01 -0800, Howard Chu wrote:
> >just fyi, I timed test008 between v2.0.18 and HEAD, doing just
> search/read
> >(no add). (Optimized, no profiling code.)
> >         v2.0.18 - 7.190 seconds
> >         HEAD    - 3.146 seconds
> >
> >With adds the difference in search performance is completely
> overshadowed by
> >the time to add/delete:
> >         v2.0.18 - 52.673 sec
> >         HEAD    - 51.605 sec
> >
> >   -- Howard Chu
> >   Chief Architect, Symas Corp.       Director, Highland Sun
> >   http://www.symas.com               http://highlandsun.com/hyc
> >   Symas: Premier OpenSource Development and Support
>