AUTO_CKPTS=Off when started and no checkpoints at all during a day???
Superboer
something i do not understand:
AUTO_CKPTS are switched to on in $ONCONFIG according to the manual
AUTO_CKPTS will decide
then when to checkpoint. AUTO_CKPTS are switched off when the engine
is started see below.
So CKPTINTVL 300 # Check point interval (in sec)
in onconfig will decide when to checkpoint ( or plog 75 full or llog
wrap or..)
and i still do not see a checkpoint the whole day.
for what it is worth i have not tested yet how fast the engine can
rollforward 8 LLogs of 20 MB each
at the end of the day, i assume it is more then RTO_SERVER_RESTART=300
seconds
What am i missing??
please advice.
Superboer.
02:41:13 Performance Advisory: The physical log is too small for
automatic checkpoints.
02:41:13 Results: Automatic checkpoints are disabled.
02:41:13 Action: To enable automatic checkpoints, increase the
physical log to at least 5625480 KB.
linux
onstat -g ckp
IBM Informix Dynamic Server Version 11.10.FC2W1 -- On-Line -- Up
10:18:20 -- 6501192 Kbytes
AUTO_CKPTS=Off RTO_SERVER_RESTART=300 seconds Estimated recovery
time 21 seconds
Critical Sections Physical Log Logical Log
Clock Total Flush Block
# Ckpt Wait Long # Dirty Dskflu Total Avg Total
Avg
Interval Time Trigger LSN Time Time Time
Waits Time Time Time Buffers /Sec Pages /Sec Pages /
Sec
48881 01:48:40 Startup 8501:0xe9b018 0.0 0.0 0.0
0 0.0 0.0 0.0 4 4 0 0 1 0
48882 02:01:06 Plog 8501:0x17c4018 4.9 4.9 0.0
1 0.0 0.0 0.0 89932 18519 91359 123 2345 3
48883 02:11:39 Plog 8503:0x12d71c0 83.8 83.6 0.0
1 0.0 0.1 0.1 283203 3388 287720 520 19236 34
48884 02:30:20 Plog 8505:0x19fd064 112.8 112.6 0.0
1 0.0 0.2 0.2 226184 2009 229772 210 22342 20
48885 05:00:54 *Backup 8507:0x1021018 48.9 48.8 0.0
0 0.0 0.0 0.0 237588 4866 241277 26 17476 1
Max Plog Max Llog Max Dskflush Avg Dskflush Avg
Dirty Blocked
pages/sec pages/sec Time pages/sec pages/
sec Time
19005 741 113 5756
172 0
The physical log size is smaller than the recommended size for a
server
configured with RTO_SERVER_RESTART. Fast recovery performance might
not
be optimal. For best fast recovery performance when RTO_SERVER_RESTART
is enabled, increase the physical log size to at least 4400000 KB.
For servers configured with a large buffer pool, this might
not be necessary. See the Administrator's Guide for more information.
Based on the current workload, the physical log might be too small
to accommodate the time it takes to flush the buffer pool during
checkpoint processing. The server might block transactions during
checkpoints.
If the server blocks transactions, increase the physical log size to
at least 5625480 KB.
IBM Informix Dynamic Server Version 11.10.FC2W1 -- On-Line -- Up
10:22:40 -- 6501192 Kbytes
Physical Logging
Buffer bufused bufsize numpages numwrits pages/io
P-1 45 64 882242 14012 62.96
phybegin physize phypos phyused %used
1:263 500000 105316 46125 9.22
Logical Logging
Buffer bufused bufsize numrecs numpages numwrits recs/pages
pages/io
L-2 0 64 1903959 86983 4303 21.9
20.2
Subsystem numrecs Log Space used
OLDRSAM 1903954 171681180
HA 5 180
Art Kagel
What do you mean by "rollforward 8 LLogs of 20 MB each at the end of the day"? Logical log rollforward will only happen if there is a crash and there are incomplete transactions after the last checkpoint at restart time. Are you saying that there have been 8 logical log changes since the last checkpoint and it has been all day?
Art
Art S. Kagel
Advanced DataTools (www.advancedatatools.com)
Blog: http://informix-myview.blogspot.com/
Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on my employer, Advanced DataTools, the IIUG, nor any other organization with which I am associated either explicitly, implicitly, or by inference. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves.
_______________________________________________
Informix-list mailing list
Inform...@iiug.org
http://www.iiug.org/mailman/listinfo/informix-list
Superboer
first of all thanks for the responce.
at the end of the day there is 8 logical logs ( 160 MB ) of
transactions done
so there is logable work done.
See you
Superboer.
On 16 aug, 19:56, Art Kagel <art.ka...@gmail.com> wrote:
> If there was no update/delete/insert/DDL activity on the server, then when a
> checkpoint starts it decides to exit without doing anything. Was there any
> loggable activity on the server during the day?
>
> What do you mean by "rollforward 8 LLogs of 20 MB each at the end of the
> day"? Logical log rollforward will only happen if there is a crash and
> there are incomplete transactions after the last checkpoint at restart
> time. Are you saying that there have been 8 logical log changes since the
> last checkpoint and it has been all day?
>
> Art
>
> Art S. Kagel
> Advanced DataTools (www.advancedatatools.com)
> Blog:http://informix-myview.blogspot.com/
>
> Disclaimer: Please keep in mind that my own opinions are my own opinions and
> do not reflect on my employer, Advanced DataTools, the IIUG, nor any other
> organization with which I am associated either explicitly, implicitly, or by
> inference. Neither do those opinions reflect those of other individuals
> affiliated with any entity with which I am affiliated nor those of the
> entities themselves.
>
> > Informix-l...@iiug.org
> >http://www.iiug.org/mailman/listinfo/informix-list
jrenaut
If you use RTO_SERVER_RESTART then it controls how frequently you
checkpoint and CKPTINTVL is ignored. From your onstat output it
thinks that it can roll forward through the 8 logs in 21 seconds and
in this case it won't do a checkpoint until it think based on the
RTO_SERVER_RESTART parameter it needs to do one so it can come back up
in the time specified. I'm not sure how accurate that is, but it's
based on the RAS_PLOG_SPEED and RAS_LLOG_SPEED which the server
modifies each time it get a chance to go through recovery, so over
time it gets more and more accurate in estimating how fast it can go
through recovery on your system so that it can do a better job of
making sure it meets the RTO_SERVER_RESTART time.
Jacques Renaut
IBM Informix Advanced Support
APD Team
Art Kagel
Art
Art S. Kagel
Advanced DataTools (www.advancedatatools.com)
Blog: http://informix-myview.blogspot.com/
Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on my employer, Advanced DataTools, the IIUG, nor any other organization with which I am associated either explicitly, implicitly, or by inference. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves.
Superboer
Thanks!! found it; as Jacques said:
When you use the RTO_SERVER_RESTART configuration parameter:
– The database server ignores the CKPTINTVL configuration parameter.
so i should read the manual better.
thanks again for your comments
Superboer
On 16 aug, 21:09, Art Kagel <art.ka...@gmail.com> wrote:
> Then I agree that you should be seeing checkpoints. I'm assuming no
> messages in the online log, does onstat -g ckp report any recent checkpoint
> activity? Time to call IBM support I think.
>
> Art
>
> Art S. Kagel
> Advanced DataTools (www.advancedatatools.com)
> Blog:http://informix-myview.blogspot.com/
>
> Disclaimer: Please keep in mind that my own opinions are my own opinions and
> do not reflect on my employer, Advanced DataTools, the IIUG, nor any other
> organization with which I am associated either explicitly, implicitly, or by
> inference. Neither do those opinions reflect those of other individuals
> affiliated with any entity with which I am affiliated nor those of the
> entities themselves.
>
> > Informix-l...@iiug.org
> >http://www.iiug.org/mailman/listinfo/informix-list
jrenaut
> Then I agree that you should be seeing checkpoints. I'm assuming no
> messages in the online log, does onstat -g ckp report any recent checkpoint
> activity? Time to call IBM support I think.
>
> Art
>
> Art S. Kagel
> Advanced DataTools (www.advancedatatools.com)
> Blog:http://informix-myview.blogspot.com/
>
> Disclaimer: Please keep in mind that my own opinions are my own opinions and
> do not reflect on my employer, Advanced DataTools, the IIUG, nor any other
> organization with which I am associated either explicitly, implicitly, or by
> inference. Neither do those opinions reflect those of other individuals
> affiliated with any entity with which I am affiliated nor those of the
> entities themselves.
>
> >http://www.iiug.org/mailman/listinfo/informix-list
If he has RTO_SERVER_RESTART, then no he shouldn't see checkpoints
until the server thinks it needs to do one to be able to make the
RTO_SERVER_RESTART time. From the Online Administrators Guide:
The database server provides two methods for determining how long fast
recovery
will take in the event of an unplanned outage.
v Use the CKPTINTVL configuration parameter to specify how frequently
the
server triggers checkpoints.
v Use the RTO_SERVER_RESTART configuration parameter to specify how
long
fast recovery should take.
When you use the RTO_SERVER_RESTART configuration parameter:
– The database server ignores the CKPTINTVL configuration parameter.
– The database server monitors the physical and logical log usage to
estimate
the duration of fast recovery. If the server estimates that fast
recovery will
exceed the time specified in the RTO_SERVER_RESTART configuration
parameter, the server automatically triggers a checkpoint.