Re: Thoughts on Logical Log use requested
Posted in 2006
Topics: Storage & Space Management, Logging & Checkpoints, Platform-Specific Issues, Cloud, Docker & Containers
Just to add some more info, the logs are backed up onto Netbackup managed disk and then copied to tape by a background process. Log backup time is about 25 seconds. I get alerted if more than 5% of logs need a backup so I should never get into the situation where I run out of available logs Regards Colin There are 10 types of people in the world, those that understand binary and those that don't >From: Martin Fuerderer <MARTINFU@de.ibm.com> >To: "Colin Dawson" <cjd_1955@hotmail.com> >CC: informix-list@iiug.org, informix-list-bounces@iiug.org >Subject: Re: Thoughts on Logical Log use requested >Date: Mon, 13 Mar 2006 16:18:11 +0100 > >Hi, > >that is no problem. > >Assuming you do backup your logs the only thing you need to >be concerned with is that the backup of the logs can keep up >with the logs filling. If log backup falls too much behind (e.g. >40 or 60 log files behind), then there is a chance that during >a high peak activity phase you'd run into a logs-full situation. > >Regards, >Martin >-- >Martin Fuerderer >IBM Informix Development Munich, Germany >Information Management > >informix-list-bounces@iiug.org wrote on 13.03.2006 14:16:15: > > Informix 7.31.FD7 > > Solaris 9 > > Logical Logs 80 * 100Mb, spread across 4 dbspaces, consecutive logs do >not > > reside in the same dbspace. > > > > > > During busy periods the logical logs on my production system can fill in > > > minutes sometime this is can be < 1 minute > > > > Does this matter? > > What effect is this likely to have on the DB? > > > > My own thoughts are that it is no big deal except that at the time a log > > > switch occurs I would get a small (milliseconds?) delay. I rarely see >any > > threads waiting for log buffers. > > > > Regards > > > > Colin > > > > There are 10 types of people in the world, those that understand binary >and > > those that don't > > _______________________________________________ > > Informix-list mailing list > > Informix-list@iiug.org > > http://www.iiug.org/mailman/listinfo/informix-list >
Colin Dawson wrote:
> Just to add some more info, the logs are backed up onto Netbackup
> managed disk and then copied to tape by a background process. Log backup
> time is about 25 seconds. I get alerted if more than 5% of logs need a
> backup so I should never get into the situation where I run out of
> available logs
>
>
>
> Regards
>
> Colin
>
> There are 10 types of people in the world, those that understand binary
> and those that don't
>
>
>
>
>
>> From: Martin Fuerderer <MARTINFU@de.ibm.com>
>> To: "Colin Dawson" <cjd_1955@hotmail.com>
>> CC: informix-list@iiug.org, informix-list-bounces@iiug.org
>> Subject: Re: Thoughts on Logical Log use requested
>> Date: Mon, 13 Mar 2006 16:18:11 +0100
>>
>> Hi,
>>
>> that is no problem.
>>
>> Assuming you do backup your logs the only thing you need to
>> be concerned with is that the backup of the logs can keep up
>> with the logs filling. If log backup falls too much behind (e.g.
>> 40 or 60 log files behind), then there is a chance that during
>> a high peak activity phase you'd run into a logs-full situation.
>>
>> Regards,
>> Martin
>> --
>> Martin Fuerderer
>> IBM Informix Development Munich, Germany
>> Information Management
>>
>> informix-list-bounces@iiug.org wrote on 13.03.2006 14:16:15:
>> > Informix 7.31.FD7
>> > Solaris 9
>> > Logical Logs 80 * 100Mb, spread across 4 dbspaces, consecutive logs do
>> not
>> > reside in the same dbspace.
>> >
>> >
>> > During busy periods the logical logs on my production system can
>> fill in
>>
>> > minutes sometime this is can be < 1 minute
>> >
>> > Does this matter?
>> > What effect is this likely to have on the DB?
>> >
>> > My own thoughts are that it is no big deal except that at the time a
>> log
>>
>> > switch occurs I would get a small (milliseconds?) delay. I rarely see
>> any
>> > threads waiting for log buffers.
>> >
>> > Regards
>> >
>> > Colin
>> >
>> > There are 10 types of people in the world, those that understand binary
>> and
>> > those that don't
>> > _______________________________________________
>> > Informix-list mailing list
>> > Informix-list@iiug.org
>> > http://www.iiug.org/mailman/listinfo/informix-list
>>
>
>
Humm,
An onstat -l (just the first bit) would be useful.
Seems like quite a high log usage.
Do you use unbuffered logging?
How big is your LOGBUFF??
TBP said:
>>
> Humm,
>
> An onstat -l (just the first bit) would be useful.
>
> Seems like quite a high log usage.
>
> Do you use unbuffered logging?
>
> How big is your LOGBUFF??
Darn. Beat me to it. :o|
--
Bye now,
Obnoxio
"It's easier with pictures."
-- Cosmo
"But wait, it gets worse."
-- Cosmo
"Run, don't walk, for the nearest exit."
-- Cosmo
It's a very active system.
LOGBUFF is 64, unbuffered logging is used. Buffered logging is not an
option.
Regards
Colin
There are 10 types of people in the world, those that understand binary and
those that don't
>From: TBP <TheBigPotato@NotHere.Co.Uk>
>To: informix-list@iiug.org
>Subject: Re: Thoughts on Logical Log use requested
>Date: Mon, 13 Mar 2006 18:46:06 GMT
>
>Colin Dawson wrote:
> > Just to add some more info, the logs are backed up onto Netbackup
> > managed disk and then copied to tape by a background process. Log backup
> > time is about 25 seconds. I get alerted if more than 5% of logs need a
> > backup so I should never get into the situation where I run out of
> > available logs
> >
> >
> >
> > Regards
> >
> > Colin
> >
> > There are 10 types of people in the world, those that understand binary
> > and those that don't
> >
> >
> >
> >
> >
> >> From: Martin Fuerderer <MARTINFU@de.ibm.com>
> >> To: "Colin Dawson" <cjd_1955@hotmail.com>
> >> CC: informix-list@iiug.org, informix-list-bounces@iiug.org
> >> Subject: Re: Thoughts on Logical Log use requested
> >> Date: Mon, 13 Mar 2006 16:18:11 +0100
> >>
> >> Hi,
> >>
> >> that is no problem.
> >>
> >> Assuming you do backup your logs the only thing you need to
> >> be concerned with is that the backup of the logs can keep up
> >> with the logs filling. If log backup falls too much behind (e.g.
> >> 40 or 60 log files behind), then there is a chance that during
> >> a high peak activity phase you'd run into a logs-full situation.
> >>
> >> Regards,
> >> Martin
> >> --
> >> Martin Fuerderer
> >> IBM Informix Development Munich, Germany
> >> Information Management
> >>
> >> informix-list-bounces@iiug.org wrote on 13.03.2006 14:16:15:
> >> > Informix 7.31.FD7
> >> > Solaris 9
> >> > Logical Logs 80 * 100Mb, spread across 4 dbspaces, consecutive logs
>do
> >> not
> >> > reside in the same dbspace.
> >> >
> >> >
> >> > During busy periods the logical logs on my production system can
> >> fill in
> >>
> >> > minutes sometime this is can be < 1 minute
> >> >
> >> > Does this matter?
> >> > What effect is this likely to have on the DB?
> >> >
> >> > My own thoughts are that it is no big deal except that at the time a
> >> log
> >>
> >> > switch occurs I would get a small (milliseconds?) delay. I rarely see
> >> any
> >> > threads waiting for log buffers.
> >> >
> >> > Regards
> >> >
> >> > Colin
> >> >
> >> > There are 10 types of people in the world, those that understand
>binary
> >> and
> >> > those that don't
> >> > _______________________________________________
> >> > Informix-list mailing list
> >> > Informix-list@iiug.org
> >> > http://www.iiug.org/mailman/listinfo/informix-list
> >>
> >
> >
>Humm,
>
>An onstat -l (just the first bit) would be useful.
>
>Seems like quite a high log usage.
>
>Do you use unbuffered logging?
>
>How big is your LOGBUFF??
>_______________________________________________
>Informix-list mailing list
>Informix-list@iiug.org
>http://www.iiug.org/mailman/listinfo/informix-list