Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations strongm on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Disksuite Naming Convention

Status
Not open for further replies.

naggiman

Technical User
May 7, 2003
36
GB
I was wondering what the general consensus of disksuite naming convention was.

We generally use
d0 = /, mirrored over d10 and d20 where d10=c0t0d0s0 and d20=c1t0d0s0
d1 = swap, mirrored over d11 and d21 where d11=c0t0t0s1 and d21=c1t0d0s1
etc...

This works very well for only 2 disks in a system, but falls over in 4 disks.

What conventions do you use in 4 disk configs?
 
Given the limits of what you can name things, I've found all of my attempts to be crytic at best. But I use the following:

d0 = d100, d101 = c0t0d0s0, c1t0d0s0
d1 = d11, d12 = c0t0d0s1, c1t0d0s1
d3 = d31, d32 = c0t0d0s3, c1t0d0s3

I try to match the first number of the submirrors to the mirror device and the disk slice. As you can see, you have to get a lil creative with d0 but as long as the people who have to work on the system with you can follow it you'll be ok. But to be honest, there's nothing wrong with your current scheme as long as you document it.

BTW - Make sure you keep a copy of the final config in an md.tab file for safe keep using the metastat -p command. This will give you the info you need to rebuild it in case of emergencies.

Herm
 
I prefere to start with d1 instead if d0, then I changed

> cat /kernel/drv/md.conf
#
#pragma ident "@(#)md.conf 2.1 00/07/07 SMI"
#
# Copyright (c) 1992-1999 by Sun Microsystems, Inc.
# All rights reserved.
#
#fp name="md" parent="pseudo" nmd=128 md_nsets=4;
name="md" parent="pseudo" nmd=2048 md_nsets=4;

rebootet with -r option and I setup metadevices the following scheme (which is quite similar to Herm's but isn't limited to 9 metadevices - Just increase the nmd to what you need):
d1 -m d101 d102
d10 -m d1001 d1002

Regards
-- Franz
Sorry I'm not a native spaeker, I'm from Munich, Germany - "Home of the Whopper", oh no, "Home of the Oktoberfest" ;-)
Solaris System Manager; I used to work for Sun Microsystems Support (EMEA) for 5 years
 
Thanks both for your suggestions.

Franz,

Don't you find having so many devices set up for disksuite an issue? This is one thing i was shying away from (I did increase mine up to 256 so i could play with the config a bit). Are there any issues with having so many?

Thanks

Duncs
 
Duncs,

i do not know any problems resulting from having "too many" ;-) metadevices; afaik this is just an allocated array in the driver, which is just n times larger than default...

Regards
-- Franz
Sorry I'm not a native spaeker, I'm from Munich, Germany - "Home of the Whopper", oh no, "Home of the Oktoberfest" ;-)
Solaris System Manager; I used to work for Sun Microsystems Support (EMEA) for 5 years
 
And the last question then...

Is there any way to get all the extra defvices created without a reboot? devfsadm didnt seem to create them when i tried that...

Or it is a case of the kernel driver for disksuite creates them when they are loaded in, and there is no way to force create them?

Duncs
 
I rebooted my machine, this took less time than searching... ;-)

and (from sunsolve.sun.com)
Document ID: 2085
SYNOPSIS: Ten common Solstice Disksuite questions
says

9. How do I create more than 128 meta devices?
Edit the file /kernel/drv/md.conf file and change the "nmd" field from
nmd=128 to whatever amount is needed (i.e., 255).
Then boot -r to build these new devices.


I guess they are created when the driver is loaded...





Regards
-- Franz
Sorry I'm not a native spaeker, I'm from Munich, Germany - "Home of the Whopper", oh no, "Home of the Oktoberfest" ;-)
Solaris System Manager; I used to work for Sun Microsystems Support (EMEA) for 5 years
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top