Re: checkpoint duration
Hi Willie,
Informix will perform a checkpoint either after CKPTINTVL seconds,
( in your case every 8 hours - that's not the case ),
or when the physical log file fills up to 75%. I guess this is
the case.
Your checkpoint occurs every 2 minutes, while you spend appr. 1 min.
for the checkpoint itself. So someone is inserting/deleting or
updating a lot of rows. Let's say, 56MB per minute.
I guess your disks are slow or you read a lot of data concurrently
from the same disks. Possibly you have a buggy version, like
Art S. Kagel told you.
I think it's a good idea to tell us more about your disks and
whether you are using KAIO or the classic I/O. (I guess you are
using KAIO). I'm also interested in your number of chunks.
If you have enough time, read also this page:
http://www.weideneder.de/informix/faq/checkdur.html
Bye
Stefan Weideneder
> NOTE: This is a non-logging database for decision support. The long
> checkpoints occur during posting of data.
>
> The following is a listing of the parameters from the onconfig file.
>
> PHYSFILE 75000 # Physical log file size (Kbytes)
> LOCKS 200000 # Maximum number of locks
> BUFFERS 200000 # Maximum number of shared
> buffers(12/11/97)
> NUMAIOVPS 1 # Number of IO vps (tap was 0 3/5/97)
> PHYSBUFF 512 # Physical log buffer size (Kbytes)
> LOGBUFF 32 # Logical log buffer size (Kbytes)> LOGSMAX 40 # Maximum number of logical log files
> CLEANERS 120 # Number of buffer cleaner
> processes(12/11/97)
> SHMBASE 0x10000000 # Shared memory base address
> SHMVIRTSIZE 1280000 # initial virtual shared memory segment> size
> SHMADD 128000 # Size of new shared memory segments
> (Kbytes)
> SHMTOTAL 1900000
> CKPTINTVL 30000 # Check point interval (in sec)
> LRUS 64 # Number of LRU queues(2/20/98)
> LRU_MAX_DIRTY 2 # LRU percent dirty begin cleaning limit
> LRU_MIN_DIRTY 1 # LRU percent dirty end cleaning limit>
> Regards,
>
> Willie Meeks