Re: disk naming proposal & devfs (fwd)

Jauder Ho (jauderho@transmeta.com)
Mon, 12 Jan 1998 23:05:58 -0800 (PST)


leonard's naming scheme sounds reasonable.. I would have preferred the
names to be a little closer to what solaris defines and just throw in the lun
as l but that's okay.

--jauder

On Tue, 13 Jan 1998, Richard Gooch wrote:

> Perry Harrington writes:
> > Greetings, I am new to (this) list. I have quite a bit of experience with
> > Solaris, and I must say that duplicating the Solaris naming scheme to access
> > SCSI disks/devices is a Good Thing(tm). Getting commercial software vendors
> > to support Linux is difficult already, creating some non-standard obscure
> > naming scheme is a step in the wrong direction. It is conceivable that we
> > could get some commercial vendor (Veritas???) to support Linux when logical
> > vols are up to snuff, and this would make them happy. I personally think
> > that the Controller,Target,device?,slice model seems kindof silly, but Solaris
>
> There is no way we can use the Solaris format, since they miss
> information (the SCSI channel or bus, which is a per host/controller
> thing). Leonard has come up with a reasonable alternative to what I've
> done: h0c0t0u0p2: host,channel,target,unit,partition.
>
> Regards,
>
> Richard....
>

Tag. You're IT.