Re: Relation extension scalability

On 10/03/16 09:57, Dilip Kumar wrote:
>
> On Thu, Mar 10, 2016 at 7:55 AM, Petr Jelinek <petr@2ndquadrant.com
> <mailto:petr@2ndquadrant.com>> wrote:
>
> Thanks for the comments..
>
>     Hmm, why did you remove the comment above the call to
>     UnlockRelationForExtension?
>
> While re factoring I lose this comment.. Fixed it
>
>     It still seems relevant, maybe with some minor modification?
>
>     Also there is a bit of whitespace mess inside the conditional lock
>     block.
>
> Fixed
>
> I got the result of 10 mins run so posting it..
> Note: Base code results are copied from up thread...
>
> Results For 10 Mins run of COPY 10000 records of size 4 bytes script and
> configuration are same as used in up thread
> --------------------------------------------------------------------------------------------
>
> Client    Base    Patch
> 1            105    111
> 2            217    246
> 4            210    428
> 8            166    653
> 16          145    808
> 32          124    988
> 64            ---    974
>
>
> Results For 10 Mins run of INSERT 1000 records of size 1024 bytes(data
> don't fits in shared buffer)
> --------------------------------------------------------------------------------------------------
>
> Client    Base    Patch
> 1            117    120
> 2            111    126
> 4             51     130
> 8             43     147
> 16           40     209
> 32           ---      254
> 64           ---      205
>

Those look good. The patch looks good in general now. I am bit scared by 
the lockWaiters * 20 as it can result in relatively big changes in rare 
corner cases when for example a lot of backends were waiting for lock on 
relation and suddenly all try to extend it. I wonder if we should clamp 
it to something sane (although what's sane today might be small in 
couple of years).

--   Petr Jelinek                  http://www.2ndQuadrant.com/  PostgreSQL Development, 24x7 Support, Training &
Services