clearing Vinum configurations

[aarong]

Firstly, let me extend my utmost thanks for helping me these past few 
weeks, Mr. Lehey. Staring down the barrel of a Political Science major, 
your coding has made me rethink my decision to steer away from a 
Computer Science degree.

On Tuesday, October 14, 2003, at 10:04  PM, Greg 'groggy' Lehey wrote:

>
> I'm going to need more information on this.  It doesn't correspond at
> all with my experience.  Here's an example:
> ...

I found my problem, I would obliterate a configuration and not 
saveconfig. It is my experience that resetconfig will obliterate only 
the configuration currently read into memory, and NOT clear the Vinum 
configuration written to Vinum disks - in my case da0 and da1. If I 
omitted saving the config after resetting, Vinum would read the 
previous configuration (after reboot) on the drives and start all over 
again. I have conquered this problem only to come to another perplexing 
Vinum problem.

On another box, again 4.8-RELEASE only with IDE instead of SCSI drives, 
I completed setup and began to mirror. Without thinking I tried to add 
plexes on an uninitialized drive and as such, Vinum read the new device 
name beta, but listed the drive in a state of referenced. I initialized 
the drive per Handbook instructions, copied the disklabel from ad0 to 
ad3 (drive beta) and tried again. After successfully starting each new 
plex, I saved the config, double checked the config written to both 
disks with dumpconfig, and rebooted. Upon reboot I was back to square 
one. I've replicated this three times now. Did you know it takes 2 
hours two completely mirror a 69GB slice across two IDE drives? ;-)

What I find VERY interesting is that ad3, hence known to vinum as 
"beta", has the correct dumpconfig output: all plexes are "state up" 
and everything is kosher. Drive "alpha", aka ad0, has the incorrect 
original config where I accidently tried to create plexes on an 
uninitialized drive. All *.p1 plexes are "state faulty" on "alpha" and 
all the saveconfig's in the world don't seem to be changing this. I, of 
course, have diligently searched the mailing lists looking for anyone 
with a similar problem, only to find another guy who didn't initialize 
the second drive correctly but then promptly fixed it.

Why would the two drives have differing configurations? Why can I 
remove all the faulty plexes, subdisks, and the "referenced" state beta 
drive; re-add beta and create new plexes; start them; save the config; 
have everything run fine; only to lose it all after rebooting?

Once again I retreat to my reference box here next to me for more 
testing tomorrow.

>
>> 2) why doesn't dd if=/dev/zero of=/dev/da0 seek=16 count=265 work?
>> What's in those 265 sectors that I can't touch?
>
> The second sector on the disk is protected.  You can't write to it.
> If you start at sector 2, it will work.
>
> The Vinum label is at sector 8.  That's the only sector you need to
> clear to manually obliterate a Vinum configuration.

Then the command should look something like: dd if=/dev/zero 
of=/dev/da0 seek=18 count=263 ? Sector eight after the sixteen sector 
bootstrap, correct?

Again, my utmost thanks.
-aarong

>
> Greg
> --
> When replying to this message, please copy the original recipients.
> If you don't, I may ignore the reply or reply to the original 
> recipients.
> For more information, see http://www.lemis.com/questions.html
> See complete headers for address and phone numbers.
> <mime-attachment>