|
Hi!
In multi-user environments IO times won't probably
be much slower with 1M extensions, because disk reading heads will very likely
move away from original location right after first request to serve other
sessions request. Also, if you think about indexed access to tables, then data
from completely different locations of Oracle datafiles are read
anyway.
You just should make sure that your extent sizes
match the (multiple of) size of db_block_size * db_file_multiblock_read_count.
That way if you even got 32 block size * 64 d_f_m_r_c = 2MB and autoextend 1M
you won't have any problem, because if extent size is 2MB, 2MB is autoextended
anyway.
I personally prefer to precreate datafiles (usually
with uniform sizes) for highly loaded systems and monitor space in them, for the
reason that otherwise lot's of extra IO is generated when extending 100M for
example. Also, the session extending segment, has to wait until the extension is
complete... This might cause locking explosions etc.. in high-concurrency
systems etc.. Instead, I monitor the free space inside tablespaces and add
datafiles during low activity periods.
But for small databases in sense of disk
usage and load, I usually use autoextend and monitor free space on disk
just to avoid extra maintenance.
Tanel.
|
- Increase tablespace, which way is better? Liu, Jack
- Re: Increase tablespace, which way is better? AK
- RE: Increase tablespace, which way is better? Stephane Paquette
- RE: Increase tablespace, which way is better? David Wagoner
- RE: Increase tablespace, which way is better? Liu, Jack
- Re: Increase tablespace, which way is better? Jos
- RE: Increase tablespace, which way is better? Jamadagni, Rajendra
- Re: Increase tablespace, which way is better? Tanel Poder
- Re: Increase tablespace, which way is better? Tanel Poder
