> Does the 4M ginsu-knife approach buy you something? Yes, it allows you
> to have infinitely configurable partitions, which can be scattered
> across the entire disk in a non-contiguous fashion. Whether or not this
> is a good thing or not can be debated. I will say that if the
> filesystem isn't involved, some of its optimizations to reduce seek
> times get thrown out the door since there are no guarantees whether
> adjacent 4M blocks are anywhere near each other or not. Then again,
> some people may prefer the ability to create partitions without needing
> any kind of advance planning as being more important than performance.
> (I don't, but clearly some people do.)
The main issue here is that the chunks of 4MB (or larger) are the basic
unit they use for mirroring updates and stale mirror area marking. It
believe it has reasons there. And you also get hit by the performance
penalty of random partition layout only if you need the flexibility. As
soon as things settle down after having grown filesystems you can run
reorgvg which will make everything contiguous again. Reorgvg can be run
while the system is up and filesystems are mounted, but due to the disk
bandwidth consumed you should probably not run it at the peak hours.
I run several AIX servers daily and I really enjoy the flexibility it
gives me with the disks. I have multiple times used the migratepv command
to move data from disks that started to get flakey to good disks while the
file systems where in use (just needing a quick reboot to connect the
hardware safely). The migratepv command also uses mirroring in the 4MB
chunk size to get the data to the new disk and afterwards breaks the
mirror to the old disk.
-- Jens-Uwe Mager <pgp-mailto:62CFDB25>- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/