Re: Re: Moving a live production database to different server and postgres release - Mailing list pgsql-general

From Alban Hertroys
Subject Re: Re: Moving a live production database to different server and postgres release
Date
Msg-id A8C2FC10-C755-46EE-85F0-27C978B5B09D@solfertje.student.utwente.nl
Whole thread Raw
In response to Re: Re: Moving a live production database to different server and postgres release  (Guillaume Lelarge <guillaume@lelarge.info>)
List pgsql-general
On 15 Jun 2010, at 9:21, Guillaume Lelarge wrote:

> Le 15/06/2010 09:18, Ulas Albayrak a écrit :
>> Unfortunately, the switch to Windows is out of my hands. If it were up
>> to me I'd stick with BSD. When you say postgres on Windows is known
>> for its "mediocre performance", do you mean it's slower or buggy? Or
>> both?
>>
>
> Slower. If it were buggy, it would be fixed.


Yes, bugs do get fixed, but bugs that haven't been fixed yet are still bugs.

Why I'm saying this is that Postgres on Windows is relatively new compared to other operating systems. In general, the
longersoftware exists and the more users it has on a certain operating system, the more bugs get fixed. 

Now I don't know the usage numbers on the different operating systems, the number of implementations on Windows has
definitelygrown a lot in the last years, but historically the Windows version used to have a relatively small user base
andwas therefore more likely to contain bugs. 

That aside, I think Windows is the only supported OS that has API's that differ a lot from the usual API's that
Postgreswas developed for. Therefore the Windows version has its own set of potential problems. 

I'm not saying that Postgres on Windows is buggy, I'm just saying that the chances you run into one are relatively
speakinghigher on Windows than on other operating systems. If you check the archives you won't see a lot of bug reports
though,no matter what OS people are using. 

And yes, it's slower on Windows. IIRC that's because Windows isn't very good at multi-processing and Postgres runs as
multipleprocesses. 

Alban Hertroys

--
If you can't see the forest for the trees,
cut the trees and you'll see there is no forest.


!DSPAM:737,4c17598b286211489720513!



pgsql-general by date:

Previous
From: Leonardo F
Date:
Subject: Re: Partial indexes instead of partitions
Next
From: Greg Smith
Date:
Subject: Re: Re: Moving a live production database to different server and postgres release