> > 1. So from an oracle perspective ZFS makes
> updates/deletes fast but
> full table scans are >slow - this is bad :(
>
> I'm by no means a database expert. But haven't the
> full table scans
> improved in the last years? Maybe in this particular
Yes full table scans have improved greatly since they started to use extent
based filesystems which is exactly my point. WAFS (Write Anywhere FileSystems)
on the other hand will degrade performance of full table scans- ZFS is a WAFS.
> case there is a
> performance hit, where read-ahead might perform well
> enough. But then
> again, the update/delete of random data was probably
> faster... There was
So to put it simply again here we are discussing oracle on ZFS and if the
active set of your database is larger than main memory and you do a lot of
batch jobs that do full table scans then sequential read performance is
important to you and ZFS or any other WAFS is a bad choice and any extent based
filesystem is a good choice QFS ASM etc
Rgds
Robin
> an interesting thread a while ago:
> http://www.opensolaris.org/jive/thread.jspa?messageID=
> 14997
^^^
A little of topic but a good read
This message posted from opensolaris.org
_______________________________________________
opensolaris-discuss mailing list
[email protected]
