Bad idea to shrink SHARED_POOL_SIZE dynamically?

[a128]

Environment: Oracle 9.2.0.1 on Solaris 8

Is it a bad idea to shrink SHARED_POOL_SIZE from 300MB to 200M using

ALTER SYSTEM SET SHARED_POOL_SIZE=200M SCOPE=BOTH?

It's been taking 15 mins since I issued the command and I am still waiting for the return prompt.

I notice the checkpoint process has been pretty busy with consistent 50% CPU usage. Does it essentially flush the shared pool and allocate it?

BTW, I do this because I have to increase JAVA_POOL_SIZE significantly (thus reduce SHARED_POOL_SIZE temporarily) in order to accomodate the 9.2.0.4 patch's requirement for JAVA_POOL_SIZE > 150M.

Thanks

[SANJAY_G]

I guess Oracle will flush shared_pool but even then it shouldn't take that long.
I haven't yet applied 9.2.0.4 patchset.

[TimHall]

Hi.

Every time I've attempted to shrink the shared pool it's taken an absolute age. It's still alot quicker to bounce the instance if you can.

Cheers