When are checkpoints initiated
I know there are a number of things that can cause a checkpoint to be initiated. The one I'm concerned about is when a checkpoint is initiated because your <blank> is 75% full/dirty. Is <blank> buffers, logs (logical or physical), or log buffers. Any help would be appreciated!
RTFM. There's a whole section on this specific topic. Top Cat wrote in message <86akgl$omp$1@nnrp03.primenet.com>... >I know there are a number of things that can cause a checkpoint to be >initiated. The one I'm concerned about is when a checkpoint is initiated >because your <blank> is 75% full/dirty. Is <blank> buffers, logs (logical >or physical), or log buffers. Any help would be appreciated! > >
Physical log. Top Cat <cstefanick@ocsmgmt.com> wrote in message news:86akgl$omp$1@nnrp03.primenet.com... > I know there are a number of things that can cause a checkpoint to be > initiated. The one I'm concerned about is when a checkpoint is initiated > because your <blank> is 75% full/dirty. Is <blank> buffers, logs (logical > or physical), or log buffers. Any help would be appreciated! > >
In article <86akgl$omp$1@nnrp03.primenet.com>,
Top Cat <cstefanick@ocsmgmt.com> wrote:
>I know there are a number of things that can cause a checkpoint to be
>initiated. The one I'm concerned about is when a checkpoint is initiated
>because your <blank> is 75% full/dirty. Is <blank> buffers, logs (logical
>or physical), or log buffers. Any help would be appreciated!
As Neil suggested, you can read the fine (and it is indeed fine) manual
for the details.
The answer to this question is "physical log," but you can get checkpoints
based on time (CKPTINTVL) or on logical log usage (next logical log contains
last checkpoint). Archives and clean shutdowns also give you a checkpoint.
Finally, you can force a checkpoint yourself with "onmode -c".
--
W. Jim Jordan, Nortel Networks, Stop 29CA3A08 | +1 613 763 1568
PO Box 3511 Stn C, Ottawa, ON K1Y 4H7 Canada | wjjordan@nortelnetworks.com
Opinions expressed are not necessarily those of Nortel Networks.
Inbound spam filtering is in place. Don't send what I won't see.
Chekpoint is initiated when Physical log if 75% full, so consider increasing the size of physical log if chekpoints are too frequents. regards Top Cat a écrit: > I know there are a number of things that can cause a checkpoint to be > initiated. The one I'm concerned about is when a checkpoint is initiated > because your <blank> is 75% full/dirty. Is <blank> buffers, logs (logical > or physical), or log buffers. Any help would be appreciated!
A checkpoint REQUEST is raised at phylog being 75% full (among a number of other operations). We may or may not initiate a checkpoint immediately - we have to let any critical sections finish first. The real danger of the phylog being too small is physical log overflow, whereas we continue writing past the logical beginning of the phylog, or in technical terms - the dog has swallowed it's tail. IF the checkpoints completes, we're ok as we move the beginning (give the phylog a new logical beginnning). If the engine dies prior to the checkpoint completing, you'll typically hang in fast recovery when trying to bring up the engine. Fabrice Dubosc wrote: > Chekpoint is initiated when Physical log if 75% full, so consider increasing > the size of physical log if chekpoints are too frequents. > > regards > > Top Cat a écrit: > > > I know there are a number of things that can cause a checkpoint to be > > initiated. The one I'm concerned about is when a checkpoint is initiated > > because your <blank> is 75% full/dirty. Is <blank> buffers, logs (logical > > or physical), or log buffers. Any help would be appreciated!