Date: Thu, 1 Jun 2000 17:56:08 +0930 From: Greg Lehey <grog@lemis.com> To: tomb <tomb@cgf.net> Cc: darren@profero.com, FreeBSD Questions <questions@FreeBSD.org> Subject: Re: vinum and bad pack magic number Message-ID: <20000601175607.O20158@wantadilla.lemis.com> In-Reply-To: <393764F2.A0D05342@cgf.net> References: <002401bfcb06$a3ecf140$a99d24d4@vindaloo.profero.com> <39353A62.DDBAB828@cgf.net> <20000601111128.F20158@wantadilla.lemis.com> <393764F2.A0D05342@cgf.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On Friday, 2 June 2000 at 0:40:34 -0700, tomb wrote:
> Greg & Darren (off list)
Sorry, I'm going to have to put this back on the list. What you are
claiming is confusing and misleading.
>>> [claim that you must set 4.2BSD partition type]
>>
>> This is not correct. If you do that, you *cannot* use them for Vinum.
>
> Ether way you get there, it was not obvious from the documentation
> how to get operational partitions.
What you're suggesting is about as relevant as turning around three
times and throwing a pinch of salt over your left shoulder. I can't
let that kind of statement go.
> I spent a while scraching my head and discovered that vinum did
> require some kind of disklabeling that could not be achieved with
> disklabel on new disks.
This is just plain *wrong*.
> If the format was not necessary then part of what I said is
> misleading, however it worked and that is very important to me.
Did you throw any salt over your shoulder? It *didn't* *work*, but
after you did it you found the correct solution.
> By operating /stand/sysinstall it was possible to get a disklabel
> that could be edited in a meaningful way. (By the way the suggestion
> to do this came from another person having similar problems)
So does using disklabel, which is the preferred way.
> Nothing in the documentation that I read (which was very extensive
> and so I probably missed a bit as I fell asleep) had a simple
> description of what was necessary to complete the task. Although
> the documentation is a very technicaly interesting read, I was still
> left with some fundamental questions left un-answered which is no an
> uncommon feature of technical literature.
So ask those questions, or experiment. But don't describe your
experimentation as the correct way to solve the problem.
> The key point that I was searching for was that it was not possible
> for me to take a newly formated disk straight from the box and
> connect as part of a vinum array.
You apparently weren't successful, but it is possible to do exactly
that.
> I spent a long time wondering if vinum operated at very low-level on
> raw media of if some manipulation was necessary.
You could have asked that question. The answer is "no".
> My attempts to use disklabel were futile until the disk had been fed
> to the /stand/sysinstall fdisk program first. There is probably a
> command line way of achieveing the same result, but at the end of
> the day this was what worked, and therefore I have no problem
> suggesting as one approach to getting operational, even if it is
> very un-elegant.
Please stop suggesting this kind of approach.
FWIW, the command line approach (which also happens to be the easiest)
is described in the man page:
DRIVE LAYOUT CONSIDERATIONS
vinum drives are currently BSD disk partitions. They must be of type
vinum in order to avoid overwriting data used for other purposes. Use
disklabel -e to edit a partition type definition. The following display
shows a typical partition layout as shown by disklabel:
8 partitions:
# size offset fstype [fsize bsize bps/cpg]
a: 81920 344064 4.2BSD 0 0 0 # (Cyl. 240*- 297*)
b: 262144 81920 swap # (Cyl. 57*- 240*)
c: 4226725 0 unused 0 0 # (Cyl. 0 - 2955*)
e: 81920 0 4.2BSD 0 0 0 # (Cyl. 0 - 57*)
f: 1900000 425984 4.2BSD 0 0 0 # (Cyl. 297*- 1626*)
g: 1900741 2325984 vinum 0 0 0 # (Cyl. 1626*- 2955*)
In this example, partition g may be used as a vinum partition. Parti-
tions a, e and f may be used as UFS file systems or ccd partitions. Par-
tition b is a swap partition, and partition c represents the whole disk
and should not be used for any other purpose.
>> This is completely incorrect. Sure, you can go through setting any
>> kind of partition type (swap, System V, whatever), but it won't work.
>> If you change it to vinum, it will work, but the history of how you
>> got there makes no difference.
>
> Sorry but that previous paragraph is confusing to me, are you saying
> that changing the label to vinum is the right or worng thing to do.
You must set the partition type to vinum. You suggested first
changing it to 4.2BSD, which is wrong.
>>> Finally you will have to tell vinum to start the array, using the
>>> "start" command. This appears to be necessary only the first time
>>> around. From then on the array will start automatically.
>
> The box power has bounced twice and I have always found that the
> array has restarted itself with no additional intervention.
So you have "start_vinum" set in your /etc/rc.conf.
> PS Greg even though I had problems getting vinum going, I find it an
> excelent system and thank you for your extensive contribution to
> this facility.
Thanks. You're welcome.
For the sake of all others reading this: sure, it's possible to have
problems setting up Vinum, especially as the installation support is
still somewhat primitive. But if you have problems, please ask. If
you don't, and you finally get it to work anyway, at least check with
me before broadcasting a description of your experiments.
Greg
--
Finger grog@lemis.com for PGP public key
See complete headers for address and phone numbers
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20000601175607.O20158>