Re: Poor quality of multipath-tools
On Wed, Jul 05, 2006 at 04:06:54PM -0500, John Goerzen wrote:
> Now, when you bring up this system, Linux will assign *two* /dev/sdx
> devices for each RAID LUN (basically looks like a disk). At any given
> time, exactly one will be readable and useful. That is, the disk can be
> probed on both controllers, but only one path will support I/O at any
> given time.
To add further complexity: there may be more than two paths to a LUN,
which means there will be more /dev/sdX devices - typically four in a
dual-controller/dual-SAN-fabric setup. And some disk systems will accept
I/O on either controller. Then it becomes sensible to try to
load-balance across the controllers for maximum throughput.
> I understand. I'm just trying to figure out if there are interested
> parties out here to pitch in, if the LVM folks have plans for it, etc.
We used the multipath tools at work (sanger.ac.uk) and I'd like to see
the package maintained well. Let me know if I can help.