Unix Technical Forum

Expanding storage on an existing SVM mirror

This is a discussion on Expanding storage on an existing SVM mirror within the Sun Solaris Administration forums, part of the Solaris Operating System category; --> I have a question about expanding storage on an existing SVM mirror. I am running Solaris 9 on a ...


Go Back   Unix Technical Forum > Unix Operating Systems > Solaris Operating System > Sun Solaris Administration

Register FAQ Members List Calendar Search Today's Posts Mark Forums Read
  #1 (permalink)  
Old 01-16-2008, 09:42 AM
Jeffery Small
 
Posts: n/a
Default Expanding storage on an existing SVM mirror


I have a question about expanding storage on an existing SVM mirror.

I am running Solaris 9 on a SunFire V250 SPARC system. I currently have
two 73 GB drives (c0t0d0 and c0t1d0) set up as a SVM RAID 1 mirror. Each
drive is partitioned as follows:

s0 60.34 GB /
s1 4.00 GB <swap>
s3 4.97 MB <replica>
s4 4.00 GB /var
s5 4.97 MB <replica>

The mirror configuration is as follows:

Mirror d0: [/]
Submirror d10
Submirror d20
Mirror d1: [swap]
Submirror d11
Submirror d21
Mirror d4: [/var]
Submirror d14
Submirror d24

I now have two additional 73 GB drives (c0t2d0 and c0t3d0). I intend to
partition each of these disks with one large s0 partition which includes
all of the available disk space. I would then like to add each as a
concatenation to the root slice on the existing mirrored drives. After
reading the SVM documentation and the various manual pages, I see example
#5 in the metattach(1m) manual page which seems to be stating that the new
disks can be added to the live mirror with the commands:

metattach d10 c0t2d0s0
metattach d20 c0t3d0s0

after which, the space is then expanded using:

growfs -M / /dev/md/rdsk/d0

Questions:

1: Is it really this simple? At first I thought I would have to break
the mirror, add the concatenations to d10 and d20 using metainit(1m),
use growfs(1m) on each of the submirrors and then rebuild the mirror.
However, since I cannot unmount the / or /var filesystems, this
procedure was also unclear. I'm also a bit puzzled how, on a live
system, you could concatenate/grow space on one of the mirrors while
the other one had not yet been expanded. Is SVM simply smart enough
to deal with this situation?

2: Other than partitioning the two new disks identically, is there any
other disk preparation that would be required? I assume that the
growfs(1m) will take care of expanding the filesystem across the
new disks.

3: If the metattach method above works, is it correct to use growfs on
the actual mirror d0 or does it have to be used on each of the d10
and d20 submirrors? I'm concerned about getting this right as I do
not want to clobber my existing OS.

4: Since we are messing around with the / filesystem, is there any need
to rerun metaroot(1m) or modify the /etc/vfstab entries? (I would
assume not.)

5: Is it necessary to update the /etc/lvm/md.tab file after these
revisions with:

metastat -p > /etc/lvm/md.tab

I would assume so, but I never saw any reference to this is the
examples.

6: Any other issues or steps I am missing?

Thanks for any help you can offer.

Regards,
--
Jeff

C. Jeffery Small CJSA LLC 206-232-3338
jeff@cjsa.com 7000 E Mercer Way, Mercer Island, WA 98040
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #2 (permalink)  
Old 01-16-2008, 09:42 AM
jeffhigham@gmail.com
 
Posts: n/a
Default Re: Expanding storage on an existing SVM mirror

Yes it is "reall [that] simple". Adding two more drives to / is a bit
questionable though. You may want to reconsider that. Why not just
create another mirror and mout it on /somedir?

In any case, here are the commands you need to run after partitioning
the secondary drives:

metattach d10 c0t2d0s0
metattach d20 c0t3d0s0
metattach d0
growfs -M / /dev/md/rdsk/d0

I have done this countless times and never had an issue.


Jeffery Small wrote:
> I have a question about expanding storage on an existing SVM mirror.
>
> I am running Solaris 9 on a SunFire V250 SPARC system. I currently have
> two 73 GB drives (c0t0d0 and c0t1d0) set up as a SVM RAID 1 mirror. Each
> drive is partitioned as follows:
>
> s0 60.34 GB /
> s1 4.00 GB <swap>
> s3 4.97 MB <replica>
> s4 4.00 GB /var
> s5 4.97 MB <replica>
>
> The mirror configuration is as follows:
>
> Mirror d0: [/]
> Submirror d10
> Submirror d20
> Mirror d1: [swap]
> Submirror d11
> Submirror d21
> Mirror d4: [/var]
> Submirror d14
> Submirror d24
>
> I now have two additional 73 GB drives (c0t2d0 and c0t3d0). I intend to
> partition each of these disks with one large s0 partition which includes
> all of the available disk space. I would then like to add each as a
> concatenation to the root slice on the existing mirrored drives. After
> reading the SVM documentation and the various manual pages, I see example
> #5 in the metattach(1m) manual page which seems to be stating that the new
> disks can be added to the live mirror with the commands:
>
> metattach d10 c0t2d0s0
> metattach d20 c0t3d0s0
>
> after which, the space is then expanded using:
>
> growfs -M / /dev/md/rdsk/d0
>
> Questions:
>
> 1: Is it really this simple? At first I thought I would have to break
> the mirror, add the concatenations to d10 and d20 using metainit(1m),
> use growfs(1m) on each of the submirrors and then rebuild the mirror.
> However, since I cannot unmount the / or /var filesystems, this
> procedure was also unclear. I'm also a bit puzzled how, on a live
> system, you could concatenate/grow space on one of the mirrors while
> the other one had not yet been expanded. Is SVM simply smart enough
> to deal with this situation?
>
> 2: Other than partitioning the two new disks identically, is there any
> other disk preparation that would be required? I assume that the
> growfs(1m) will take care of expanding the filesystem across the
> new disks.
>
> 3: If the metattach method above works, is it correct to use growfs on
> the actual mirror d0 or does it have to be used on each of the d10
> and d20 submirrors? I'm concerned about getting this right as I do
> not want to clobber my existing OS.
>
> 4: Since we are messing around with the / filesystem, is there any need
> to rerun metaroot(1m) or modify the /etc/vfstab entries? (I would
> assume not.)
>
> 5: Is it necessary to update the /etc/lvm/md.tab file after these
> revisions with:
>
> metastat -p > /etc/lvm/md.tab
>
> I would assume so, but I never saw any reference to this is the
> examples.
>
> 6: Any other issues or steps I am missing?
>
> Thanks for any help you can offer.
>
> Regards,
> --
> Jeff
>
> C. Jeffery Small CJSA LLC 206-232-3338
> jeff@cjsa.com 7000 E Mercer Way, Mercer Island, WA 98040


Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #3 (permalink)  
Old 01-16-2008, 09:42 AM
Darren Dunham
 
Posts: n/a
Default Re: Expanding storage on an existing SVM mirror

In comp.unix.solaris Jeffery Small <jeff@cjsa.com> wrote:
> Questions:


> 1: Is it really this simple? At first I thought I would have to break
> the mirror, add the concatenations to d10 and d20 using metainit(1m),
> use growfs(1m) on each of the submirrors and then rebuild the
> mirror.


Even if you couldn't do this the short way, there'd never be a reason to
do work on both sides of the submirrors. They'd be overwritten on being
attached.

You'd only want to 'growfs' the main mirror.

> However, since I cannot unmount the / or /var filesystems, this
> procedure was also unclear. I'm also a bit puzzled how, on a live
> system, you could concatenate/grow space on one of the mirrors while
> the other one had not yet been expanded. Is SVM simply smart enough
> to deal with this situation?


You expand the mirror. You're never writing directly to one side of it.

> 2: Other than partitioning the two new disks identically, is there any
> other disk preparation that would be required? I assume that the
> growfs(1m) will take care of expanding the filesystem across the
> new disks.


No. SVM takes care of expanding the metadevice across the new disks.
'growfs' knows nothing about disks. It only expands the filesystem to
make use of the entire metadevice.

> 3: If the metattach method above works, is it correct to use growfs on
> the actual mirror d0 or does it have to be used on each of the d10
> and d20 submirrors? I'm concerned about getting this right as I do
> not want to clobber my existing OS.


Only on the mirror. Never write to a submirror (or underlying
metadevice).

> 4: Since we are messing around with the / filesystem, is there any need
> to rerun metaroot(1m) or modify the /etc/vfstab entries? (I would
> assume not.)


No. But 'growfs' is not supported for the mounted root filesystem.
(That doesn't mean that it doesn't work..)

--
Darren Dunham ddunham@taos.com
Senior Technical Consultant TAOS http://www.taos.com/
Got some Dr Pepper? San Francisco, CA bay area
< This line left intentionally blank to confuse you. >
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
Reply


Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On
Forum Jump


All times are GMT. The time now is 10:58 AM.


Powered by vBulletin® Version 3.6.5
Copyright ©2000 - 2008, Jelsoft Enterprises Ltd.
Search Engine Optimization by vBSEO 3.2.0
www.UnixAdminTalk.com