Re: count(*) performance improvement ideas - Mailing list pgsql-hackers

From Stephen Denne
Subject Re: count(*) performance improvement ideas
Date
Msg-id F0238EBA67824444BC1CB4700960CB48051D625A@dmpeints002.isotach.com
Whole thread Raw
In response to Re: count(*) performance improvement ideas  ("Stephen Denne" <Stephen.Denne@datamail.co.nz>)
Responses Re: count(*) performance improvement ideas  (PFC <lists@peufeu.com>)
List pgsql-hackers
> >     The big gotcha is that these are all non-transactional 
> > : if you rollback,  
> > GD and SD stay the same, and when you issue a query, you can 
> > assume the  
> > state of SD and GD is random (due to previous queries) unless you  
> > initialize them to a known value.
> 
> Using txid_current() as a key should alleviate that.

No... hold on, it is per session, and a session can't have two or more transactions active at once can it?

(Though I hear of things called sub-transactions)

So the problem is that other functions may be using GD themselves, and your own code is at the mercy of the other
functions.Conversely you shouldn't clear GD, as some other function may be using it.
 

So you're better off using a single function for everything, and using SD within it?

There isn't any way of telling whether the function is being called for the first time in a transaction. You don't know
whento clear it.
 

Regards,
Stephen Denne.

Disclaimer:
At the Datamail Group we value team commitment, respect, achievement, customer focus, and courage. This email with any
attachmentsis confidential and may be subject to legal privilege.  If it is not intended for you please advise by reply
immediately,destroy it and do not copy, disclose or use it in any way. 
__________________________________________________________________ This email has been scanned by the DMZGlobal
BusinessQuality             Electronic Messaging Suite. 
Please see http://www.dmzglobal.com/dmzmessaging.htm for details.
__________________________________________________________________


pgsql-hackers by date:

Previous
From: Andreas 'ads' Scherbaum
Date:
Subject: Re: get rid of psql welcome message
Next
From: Bruce Momjian
Date:
Subject: Re: pg_terminate_backend() issues