Jeff Bonwick wrote:
...
Since we know that intent log blocks don't live for more than a
single transaction group (which is about five seconds), there's
no reason to allocate them space-efficiently. It would be far
better, when allocating a B-byte intent log block in an N-disk
RAID-Z group, to allocate B*N bytes but only write to one disk
(or two if you want to be paranoid). This simple change should
make synchronous I/O on N-way RAID-Z up to N times faster.
Would it make sense to keep the space-efficient allocation
code around for times when disk space gets tight (as in
less than 100 free blocks or similar) ?
Darren
_______________________________________________
zfs-discuss mailing list
zfs-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
