Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Checkpoint duration

Re: Checkpoint duration

From: hpuxrac <johnbhurley_at_sbcglobal.net>
Date: 27 Nov 2006 03:51:07 -0800
Message-ID: <1164628267.898562.150640@n67g2000cwd.googlegroups.com>

timur.gokce_at_gmail.com wrote:
> During a checkpoint the following occurs:
> - The database writer (DBWR) writes all modified database
> blocks in the buffer cache back to datafiles,
> - Log writer (LGWR) updates both the controlfile and
> the datafiles to indicate when the last checkpoint
> occurred (SCN)
>
>
> i ) If I increase or decrease # of datafiles (from 10 to 5000 or
> vise-versa) in my database ,Will this effect the time spent to
> checkpoint ?

What do you think?

> ii) By considering the time elapsed between "checkpoint started" and
> "checkpoint completed" logs in the alert.log.
> For which task above does it spend much time ?

You may want to look at various postings at asktom http://asktom.oracle.com

Tom Kyte has a new book out that you might want to consider buying.
>
>
>
> On Nov 27, 2:28 am, "hpuxrac" <johnbhur..._at_sbcglobal.net> wrote:
> > timur.go..._at_gmail.com wrote:
> > > Checkpoint process updates the file headers of the data files and
> > > flushes the dirty buffers in cache to data files.
> >
> > > i) Is there any propotional relation between "number of datafiles" and
> > > time spent to perform a checkpoint ?Often the time spent updating the file header is minor compared to
> > other checkpoint activity. Logically yes there is a relationship.
> >
> > The actual times taken will vary considerably depending on your io
> > configurations and whether you support async io etc.
> >
> > > ii) Is the majority of the time spent to checkpoint caused by "header
> > > update"(LGWR) or "flushing dirty buffers"(DBWR) ?In general it depends on the system and varies widely. What do you
> > mean by header update and LGWR?
> >
> > Log writer also goes first. If there are many dirty buffers in your
> > buffer caches then the time taken to do this flushing will probably
> > dwarf the amount of time required to update the file headers.
Received on Mon Nov 27 2006 - 05:51:07 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US