Re: poor performance revisted
In a poor-performance follow-up, Obnoxio The Clown said the LRU prioritisation bug usually shows many d0-type page records, so he was surprised Ty O'Kelly had every other symptom. Doug Agnew said the bug was long known and advised checking the Informix techinfo bug reports. Ty doubted it was the bug, since setting LRUPRIORITY made performance much worse. Unresolved.
Auto-generated by Claude from the posts below — may be imperfect; read the full thread.
From: "Ty O'Kelly" <tokelly@maxor.invalid.com>
>
>Obnoxio The Clown wrote:
>
> > > > >onstat -P | tail -5:> > > > >
> > > > >Percentages:
> > > > >Data 6.77
> > > > >Btree 92.40
> > > > >Other 0.82
> > > >
> > > > This looks ugly. What is the output of
> > > >
> > > > onstat -B | grep " d0 " | wc -l> > >
> > > 0
> >
> > Hmmmm...what's the smiley for raised eyebrows.
>
>So is there need for concern when that command returns 0? What does
>this mean?
>The FM has shed little light on the output onstat -B.
No, it's just that the LRU prioritisation bug generally has squillions of
records with a d0 page type. So I'm surprised that you seem to have all the
symptoms of it, bar this one.
________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com
In article <8m62mq$87m$1@news.xmission.com>,
"Obnoxio The Clown" <obnoxio@hotmail.com> wrote:
>
> From: "Ty O'Kelly" <tokelly@maxor.invalid.com>
> >
> >Obnoxio The Clown wrote:
> >
> > > > > >onstat -P | tail -5:> > > > > >
> > > > > >Percentages:
> > > > > >Data 6.77
> > > > > >Btree 92.40
> > > > > >Other 0.82
> > > > >
> > > > > This looks ugly. What is the output of
> > > > >
> > > > > onstat -B | grep " d0 " | wc -l> > > >
> > > > 0
> > >
> > > Hmmmm...what's the smiley for raised eyebrows.
> >
> >So is there need for concern when that command returns 0? What does
> >this mean?
> >The FM has shed little light on the output onstat -B.
>
> No, it's just that the LRU prioritisation bug generally has squillions
of
> records with a d0 page type. So I'm surprised that you seem to have
all the
> symptoms of it, bar this one.
>
How much is a squillion? Is that a million times a squirrel?
________________________________________________________________________
> Get Your Private, Free E-mail from MSN Hotmail at
http://www.hotmail.com
>
>
--
# unrm /
ksh: unrm: not found
# man cpio
Sent via Deja.com http://www.deja.com/
Before you buy.
How is it you guys get to know about all these bugs? Do you 'stumble
across' them, or is there some mega-resource that gives information
about what to look for? Or more importantly what not to do so they don't
become a problem...
In article <8m62mq$87m$1@news.xmission.com>, Obnoxio The Clown
<obnoxio@hotmail.com> writes
>
>From: "Ty O'Kelly" <tokelly@maxor.invalid.com>
>>
>>Obnoxio The Clown wrote:
>>
>> > > > >onstat -P | tail -5:>> > > > >
>> > > > >Percentages:
>> > > > >Data 6.77
>> > > > >Btree 92.40
>> > > > >Other 0.82
>> > > >
>> > > > This looks ugly. What is the output of
>> > > >
>> > > > onstat -B | grep " d0 " | wc -l>> > >
>> > > 0
>> >
>> > Hmmmm...what's the smiley for raised eyebrows.
>>
>>So is there need for concern when that command returns 0? What does
>>this mean?
>>The FM has shed little light on the output onstat -B.
>
>No, it's just that the LRU prioritisation bug generally has squillions of
>records with a d0 page type. So I'm surprised that you seem to have all the
>symptoms of it, bar this one.
>________________________________________________________________________
>Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com
>
Andrew Lennard andy@kontron.demon.co.uk
We're just 'lucky'... like the old pioneers... could always tell the ones
out front because of the arrows in their back.
This bug has been around for a while and a lot of us have been bitten by it,
so it is one of the first things I check when someone gripes about sudden
performance drop after an upgrade...
Get a techinfo center membership (Free) at www.informix.com, and you can see
the bug reports (err, excuse me, undocumented feature descriptions)...
"Andy Lennard" <andy@kontron.demon.co.uk> wrote in message
news:Bs0PhcAMeuh5Ew3H@kontron.demon.co.uk...
> How is it you guys get to know about all these bugs? Do you 'stumble
> across' them, or is there some mega-resource that gives information
> about what to look for? Or more importantly what not to do so they don't
> become a problem...
>
>
>
>
> In article <8m62mq$87m$1@news.xmission.com>, Obnoxio The Clown
> <obnoxio@hotmail.com> writes
> >
> >From: "Ty O'Kelly" <tokelly@maxor.invalid.com>
> >>
> >>Obnoxio The Clown wrote:
> >>
> >> > > > >onstat -P | tail -5:> >> > > > >
> >> > > > >Percentages:
> >> > > > >Data 6.77
> >> > > > >Btree 92.40
> >> > > > >Other 0.82
> >> > > >
> >> > > > This looks ugly. What is the output of
> >> > > >
> >> > > > onstat -B | grep " d0 " | wc -l> >> > >
> >> > > 0
> >> >
> >> > Hmmmm...what's the smiley for raised eyebrows.
> >>
> >>So is there need for concern when that command returns 0? What does
> >>this mean?
> >>The FM has shed little light on the output onstat -B.
> >
> >No, it's just that the LRU prioritisation bug generally has squillions of
> >records with a d0 page type. So I'm surprised that you seem to have all
the
> >symptoms of it, bar this one.
> >________________________________________________________________________
> >Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com
> >
>
> Andrew Lennard andy@kontron.demon.co.uk
Obnoxio The Clown wrote: > No, it's just that the LRU prioritisation bug generally has squillions of > records with a d0 page type. So I'm surprised that you seem to have all the > symptoms of it, bar this one. I question whether it is the LRU priority bug as well. I bounced our server with the environmental variable LRUPRIORITY set and the performance went to hell. Ty -- Ty O'Kelly tokelly@maxor.invalid.com