Re: Manual update

2016-12-20 Thread Marc Peters
Am 12/19/16 um 19:01 schrieb Todd Carpenter:
> Hi All,
> 
> I recently installed 6.0 and was struggling to get my softraid0 stripe to
> build properly and or boot.  I went over section 14 carefully and did some
> research.  The part that I found was missing was the creation of a 100 meg
> partition on the A slice and how to successfully create a bootable raid
> stripe. (I'm sure there is something somewhere that I missed, but I thought
> it would be awesome to include step #2 and #3 in the raid setup)

That's not needed anymore. On the page you mention, it states clearly:

"[...] and booting from softraid devices isn't supported on all of them.
It's currently only possible to boot from RAID1, RAID5 and crypto
volumes on i386, amd64 and sparc64."

The same in the manpages:
http://man.openbsd.org/?query=boot=0=8=default=OpenBSD-current

As this already works for me on several systems, i assume that something
with your process is wrong here. Or you messed up your upgrade (yes, i
upgraded some of the machines).

Hth,
Marc



Re: Manual update

2016-12-19 Thread Nick Holland
On 12/19/16 13:01, Todd Carpenter wrote:
> Hi All,
> 
> I recently installed 6.0 and was struggling to get my softraid0 stripe to
> build properly and or boot.  I went over section 14 carefully and did some
> research.  The part that I found was missing was the creation of a 100 meg
> partition on the A slice and how to successfully create a bootable raid
> stripe.

no, that's not needed.

> (I'm sure there is something somewhere that I missed, but I thought
> it would be awesome to include step #2 and #3 in the raid setup)
> 
> Target
> https://www.openbsd.org/faq/faq14.html#softraid
> Source
> http;//undeadly,org/cgi?action=article=2011xxx
   [url mangled]
that's a 2011 article, which predates bootable softraid.

> #1 - is kind of a question really.  My issue was it took me several times
> to get my raid "perfect"  during the process every failed attempt would
> result in portions of the raid remaining intact. IE fdisk would see RAID
> partitions that were not wiped or bioctl would think parts of the stripe
> were degraded etc.. ultimately my solution was to zero fill all of the
> disks and start fresh.. Is this the best way? or was there something simple
> I missed?

yeah, I found it very good to zero the first one or ten MB of any part
of the disk where a softraid partition could have been BEFORE creating
the array.

> #2  - update FAQ with the additional step to included the creation of the A
> slice as a boot device.

...[misleading info removed]...

no, not needed (unless someone broke something since I last looked)

> #3 - add the bootdevice creation steps to prepare the above boot slice

and again. no.

so.  What you propose is wrong, but you found it helped you...so sounds
like there's a problem with your process.

What did you do, what did you expect to happen, what did you see happen?

Nick.



Manual update

2016-12-19 Thread Todd Carpenter
Hi All,

I recently installed 6.0 and was struggling to get my softraid0 stripe to
build properly and or boot.  I went over section 14 carefully and did some
research.  The part that I found was missing was the creation of a 100 meg
partition on the A slice and how to successfully create a bootable raid
stripe. (I'm sure there is something somewhere that I missed, but I thought
it would be awesome to include step #2 and #3 in the raid setup)

Target
https://www.openbsd.org/faq/faq14.html#softraid
Source
http://undeadly.org/cgi?action=article=20111002154251


#1 - is kind of a question really.  My issue was it took me several times
to get my raid "perfect"  during the process every failed attempt would
result in portions of the raid remaining intact. IE fdisk would see RAID
partitions that were not wiped or bioctl would think parts of the stripe
were degraded etc.. ultimately my solution was to zero fill all of the
disks and start fresh.. Is this the best way? or was there something simple
I missed?

#2  - update FAQ with the additional step to included the creation of the A
slice as a boot device.

(I)nstall, (U)pgrade or (S)hell? s
# cd /dev
# sh MAKEDEV wd1
# fdisk -iy wd0
Writing MBR at offset 0.
# fdisk -iy wd1
Writing MBR at offset 0.
# disklabel -E wd0
Label editor (enter '?' for help at any prompt)
*> a a*
*offset: [64]*
*size: [4193216] 100m*
*Rounding to cylinder: 209600*
*FS type: [4.2BSD]*
> a d
offset: [209664]
size: [3983616]
FS type: [4.2BSD] raid
> q
Write new label?: [y]
# disklabel wd0 > protofile
# disklabel -R wd1 protofile
# bioctl -c 1 -l wd0d,wd1d softraid0
sd0 at scsibus0 targ 1 lun 0:  SCSI2 0/direct fixed
sd0: 1944MB, 512 bytes/sector, 3983088 sectors
# ^D

#3 - add the bootdevice creation steps to prepare the above boot slice

Welcome to the OpenBSD/i386 5.0 installation program.
(I)nstall, (U)pgrade or (S)hell? i
Before booting, don't forget to format the two "a" partitions and copy
kernels onto them. This can be done before or after you run the
installation script. Afterwards might be easier if you did a network
install, since the kernels will be local to you now on /mnt/:
# newfs wd0a
newfs: reduced number of fragments per cylinder group from 13096 to 13040
to enlarge last cylinder group
/dev/rwd0a: 102.3MB in 209600 sectors of 512 bytes
5 cylinder groups of 25.47MB, 1630 blocks, 3328 inodes each
super-block backups (for fsck -b #) at:
 32, 52192, 104352, 156512, 208672,
# newfs wd1a
newfs: reduced number of fragments per cylinder group from 13096 to 13040
to enlarge last cylinder group
/dev/rwd1a: 102.3MB in 209600 sectors of 512 bytes
5 cylinder groups of 25.47MB, 1630 blocks, 3328 inodes each
super-block backups (for fsck -b #) at:
 32, 52192, 104352, 156512, 208672,
# mount /dev/wd0a /mnt2
# cp /mnt/bsd* /mnt2
# umount /mnt2
# mount /dev/wd1a /mnt2
# cp /mnt/bsd* /mnt2
# umount /mnt2



Thanks,

Cheers