Re: RC1 time? - Mailing list pgsql-hackers

From Marc G. Fournier
Subject Re: RC1 time?
Date
Msg-id 20020105141127.A42799-100000@earth.hub.org
Whole thread Raw
In response to Re: RC1 time?  (Oleg Bartunov <oleg@sai.msu.su>)
List pgsql-hackers
On Sat, 5 Jan 2002, Oleg Bartunov wrote:

> On Sat, 5 Jan 2002, Marc G. Fournier wrote:
>
> > On Fri, 4 Jan 2002, Tom Lane wrote:
> >
> > > Bruce Momjian <pgman@candle.pha.pa.us> writes:
> > > >> Aside from the lwlock business, Karel seems to be seeing some problem
> > > >> in to_timestamp/to_date.
> > >
> > > > I thought Karel sent in a to_date patch yesterday that you applied.  Was
> > > > there another issue?
> > >
> > > Yes.  He reported something that looked a lot like a DST boundary
> > > problem, except it wasn't on a DST boundary date.  Thomas thought it
> > > might be a consequence of the timestamp-vs-timestamptz change from
> > > 7.1 to 7.2.  See http://fts.postgresql.org/db/mw/msg.html?mid=1345390
> > >
> > > (BTW, is anyone else noticing that fts.postgresql.org is missing an
> > > awful lot of traffic?  For example, I can't get it to show Thomas'
> > > comment on the above-mentioned thread; and that is *VERY* far from
> > > being its only omission lately.)
> >
> > We just moved it from the old server (that I have to shut down) to the new
> > one at Rackspace ... the one thing I have to do over the next short period
> > of time is to spring for a memory upgrade on that machine though, as
> > 512Meg just doesn't cut it :(
>
> I see on db.postgresql.org
>
> > vmstat -w 5
>  procs      memory      page                    disks     faults      cpu
>  r b w     avm    fre  flt  re  pi  po  fr  sr da0 da1   in   sy  cs us sy id
>  0 17 0  471224  28184  369   3   4   2 325 334   0   0  331  401 182 29  2 69
>
>  0 19 0  414556  19272  644   1   1   0 546   0   0 172  461  823 290  1  2 97
>  1 19 0  414788  23940  459   4   4   1 474 615   1 170  454  734 286  0  2 98
>  1 20 0  428592  26912  372   3  14   0 433 592   6 182  480  790 296  1  2 97
>  2 19 0  458688  30164  318   3   9   0 423 592   3 177  463  787 289  1  2 97
>  1 17 0  446848  24196  303   2   4   0 454   0   2 177  463  878 294  1  2 97
>  0 18 0  452432  29404  228   1   3   2 324 633   2 184  472  842 305  2  4 94
>  0 19 0  449724  21860  200  14   6   0 508   0   1 188  473  702 283  0  2 98
>
> disk activity is very bad, probably not balanced. I catch a moment
> when fts.postgresql.org was slow.

Most of it is due to the high swap being used .. I've had two offers so
far to help upgrade the RAM, and am looking into the costs of doing so ...




pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: Some interesting results from tweaking spinlocks
Next
From: Joe Conway
Date:
Subject: Re: pgcryto strangeness...