FW: poervault/linux query

Kevin_Burroughs@Dell.com Kevin_Burroughs at Dell.com
Mon Apr 15 18:05:01 CDT 2002

Forward to list with address correction.

-----Original Message-----
From: Burroughs, Kevin 
Sent: Monday, April 15, 2002 6:02 PM
To: 'jason andrade'
Cc: 'linux-poweredge at exchange.dell.com'
Subject: RE: poervault/linux query


I read your message and the subsequent thread with considerable interest.  I
am in Fibre Channel Storage now, but was in the Linux queue previously and
still monitor this list.

Off the top of my head, a few questions jump up:
1)  Managing a 650 with Array Manager: never heard of that being done: we
use Data Supervisor or Data Administrator
2)  Failover: we use ATF in Windows to manage failover.  Linux is not
supported in our SAN software at this time, so there is no corresponding
application.  Besides, if you only have 1 HBA, there is no second HBA to
manage failover (or the second SP for that matter).
3)  Soft peer errors are something that occurs on the 650/630, but again, it
would be highly unusual to see these in Array Manager.

The proper way to shut down the array is to take the servers down first,
then use the SPS to power off the 650 (this sends a signal to the SP telling
them to dump their cache: don't remove the 650 fan pack to shut the power
supplies down directly: this will cause data corruption).  When the 650
shuts itself down, then the cache is written and it's safe to shut down the
630's.  On power up, reverse the order, making sure the 630's are fully
initialized prior to powering up the 650 so the SP's are able to find all
their LUNs.  Once the 650 is fully initialized, then power up the servers.

Kevin Burroughs, MCSE, MCP+I
PowerVault and Fibre Channel Gold Storage Support Specialist
Dell Computer Corp. Enterprise Expert Center

-----Original Message-----
From: jason andrade [mailto:jason at dstc.edu.au]
Sent: Sunday, April 14, 2002 1:53 AM
To: Michael E Brown
Cc: linux-poweredge at exchange.dell.com
Subject: Re: poervault/linux query

On Sun, 14 Apr 2002, Michael E Brown wrote:

> Neat question. I've never done anything more than flip the power switch on
> them. Unless you have write caching turned on, WHICH YOU SHOULD NOT HAVE
> TURNED ON UNLESS YOU HAVE THE UPS OPTION, you shouldn't have to worry
> about it. Sorry about yelling, but it is an important point. (Write cache
> is turned off by default anyways. You really have to know what you are
> doing to turn it on, as there are at least two (iirc) places you need to
> touch to get it on.)

we have the SPS batteries, and dell configured the server with write
caching.  but NOT on the drives. just the array.

> > 	o get the array to do a consistency check on the luns
> > 	  as i am getting a lot of "soft peer bus error"
> Another good question. I don't know about this one. Sounds bad to me.
> Where are you getting this error message?

i see this in the openmanage array manager, when i look at the
"event log" for SPA.

> >         o migrate 2 LUNs from SPA to SPB
> There is a menu option for this in Data Administrator. (from memory) click
> on the luns and change the default owner, then tell it to switch active
> SPs. The key is to change the default owner in Data Administrator on the
> NT box. You might have to have the NT box hooked up to both SPs, though. I
> don't exactly remember. :-(

i hve tried this.  i change the default owner - i even clicked on the lun
and said "change from SPA to SPB" but nothing happens.  it says it has
changed the ownership over, but it hasn't...

i cannot hook my NT box up to both SPs.. i don't have enough HBAs to
do that :-) (only one in the NT box)

> >         o save it all
> It automatically does this.

oh good.

> >         o on the linux box, add a second HBA and then i should see
> >           2 LUNs on one HBA and 2 on the other.
> This sounds about right.

no luck so far.. 

> > another quick query on the 2450 - if you have two of them configured
> > with 2 disks in each in a mirror set and you want to "swap" disks
> > between the two, if you power down both machines and swap both
> > disks (e.g disk0 and disk1) between chassis, will that work, or
> > is there some key tied into the controller nvram on diskid etc ?
> Generally you will need to boot the machine and go into the firmware to
> tell it to ignore the nvram and re-read the config from the disk. It is
> generally pretty obvious how to do this once you get into the BIOS screen,
> but there are several knowledgeable people on the list who can guide you
> step-by-step should you need it.

in the end i chanced this on a test box.  if you move both drives over,
the bios seems to recognise that it doesn't have 2 new drives.  however
it will go into a "scrubbing" period for a while.  this didn't seem to
affect the disk/data at all, and i was able to use this on other machines
so i would give it a qualified "it does the right thing".



Linux-PowerEdge mailing list
Linux-PowerEdge at dell.com
Please read the FAQ at http://lists.us.dell.com/faq or search the list
archives at http://lists.us.dell.com/htdig/

More information about the Linux-PowerEdge mailing list