Anjan Chakraborty wrote:
>
> I am getting very high chunk writes and almost 0 LRU writes against onstat
> -F
> and
> high duration against onstat -m.
> How do I solve this problem ?
> I have CKPTINTVAL=300,NUMAIOVPS=20,CLEANERS=6,LRU_MAX_WRITE=20,
> LRU_MIN_WRITE=10,LRUS=8,BUFFERS=60000,PHYSFILE=60000,
> MULTIPROCESSOR=1,NUMCPUVPS=1,SINGLE_CPU_VP=0,RESIDENT=1
> 36 Logical log files,LOGSIZE=4000, information in 6 disks and HP-UX 10 with
> 512 M memory/2 processors and running ontape -c.
Without knowing the entire system configuration it is difficult to
make tuning suggestions. One way to force LRU writes is to reduce
the LRU_MAX_DIRTY and LRU_MIN_DIRTY to - say - 3 and 1 respectively.
This way, the work done at checkpoints is reduced and the database
"freeze" experienced during checkpoints is reduced. The chunk writes
during checkpoints are more efficient than LRU writes, so they might be
better if this is a batch application.
Hope this helps.
-Atul.