Re: Where to put RAID (was Re: LVM etc... biggest thread in history)

Matthias Urlichs (smurf@noris.de)
17 Aug 1998 07:38:42 +0200


David Lang <dlang@diginsite.com> writes:
> I am thinking ot two basic jobs.
> 1. connecting multiple chunks together into one "device" for the filesystem to
> reside on.
> 2. running multiple devices throug a blender to provide raid
> performance/failover capability.
>
> I see these a seperate enough (plus budget allowing I will take hardware raid
> solutions any day, no CPU overhead) to be done in seperate layers. [...]
Right.
>
> The question is which is worse.
> additional overhead if all features are turned on (LVM + raid managed
> seperatly)
> or
> additional code if features are not used (LVM without raid, or raid
> without LVM managed together)

Both LVM and RAID should be implemented by taking a few devices and
combining them to yield another new device (RAID), or a bunch of them
(LVM). Thus, if you use an LVM on top of a RAID array the additional
overhead is just the procedure call to the RAID's read or write code,
which should certainly be acceptable. In addition, I think the code is
maintainable a whole lot better if you keep them in separate modules.

The "is this block on the same spindle as that block" question which an
optimizing file system might want to ask its underlying device isn't really
a LVL vs. RAID problem -- both would just pass the request (which would ask
for, for instance, a device-specific struct pointer so comparisons are
meaningful) down to whichever device owns the block in question.

-- 
Matthias Urlichs
noris network GmbH

- 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.altern.org/andrebalsa/doc/lkml-faq.html