On 15 Mar 98 at 22:45, Bruce Momjian wrote:
> How are you finding the speed of the 6.3 release. People are reporting
> much better performance with it, and I can see you ran your speed test
> on the 6.3 final release.
>
The speed is much better in 6.3 final as in 6.2.1 with all patches.
But I also tested the snapshot of 2 feb 1998 (6.3) and the
performance difference between that one and the final release isn't
that big but noticable. The snapshot was faster. But in the final
release you can see some things are added such as the group by
things. The performance of that test is much much better in the final
release. Further it's now possible to compile postgres 6.3 with pgcc
1.0.1 with -O6 -mpentium which is giving a little performance gain I
think. (compiled postgres 6.3 final release with it and all test are
done on it).
For the crash-me test I have some problems with the 6.3 final
release. One bug is fixed by a patch which was mailed to me by Vadim
but one bug is still crapping the crash-me test. It's the bug when
the postgres server is crashed while you are doing a create table.
Then there is a file placed in the db dir and you can't create that
table with the same name again after a connect. So several tests
aren't run then anymore or are set to 0 because nothing can be done.
I have read in postgres-questions that there will be this week a big
patch for 6.3 and I hope it will fix the problems I have with 6.3.
I will look into your other e-mail to see why it isn't supported by
crash-me and it's in the functionlist of postgres.
Greetz...
Luuk
|-----------------------------------------------------------------------|
| Luuk de Boer -> <luuk@pi.net> |
| When you want to go fast ..... |
| Drive a cobra on the road and take a T3 line for internet ...... :-) |
| http://www.luuk.com/ -> The AC Cobra site |
|-----------------------------------------------------------------------|
