Re: Anomalies with the now() function

On 11/21/05, Michael Fuhr <mike(at)fuhr(dot)org> wrote:
> [Please copy the mailing list on replies and please don't change
> the Subject header. I've restored the original subject and am
> copying the list on this reply. I'll look at this when I get a
> chance, but maybe somebody else will answer before then.]
>
> On Mon, Nov 21, 2005 at 05:28:30PM -0000, Byrne Kevin-kbyrne01 wrote:
> > Thank you for your response and I would be intereted in your opinion on a further explanation of my problem:
> >
> > The first table table (call it Table A) gets the timestamp using timeofday - and assuming what you say below is true - Table B will take now() as start of transaction time - so I can understand that part where they could be different but considering that for moServer every operation happens in a single txn, following will be the case
> >
> > Moserver receives the event - timestamps it as 't1'
> > -- time lapse before moserver computes the transation and gives it to odbc.
> > Txn_begin- now() gets frozen to 't2'
> > Insert - now() should put it as 't2'
> > Txn_end()- done.
> >

where is the server in moserver or in another machine you reach with
the odbc connection? if there are two machines involved maybe the
first time you get the time ;) is from one machine and the second time
from the other...

just an idea...

> > So firstly t2 should always be > t1 and the difference could be a few seconds but we found sometimes t2 < t1!..
> >
> > The query is done through odbc ( I think that might be additionally causing some strange behaviour?)..
> >
> > What do you think?
>
> --
> Michael Fuhr
>
> ---------------------------(end of broadcast)---------------------------
> TIP 2: Don't 'kill -9' the postmaster
>

--
Atentamente,
Jaime Casanova
(DBA: DataBase Aniquilator ;)