AF:
NF:0
PS:10
SRH:1
SFN:
DSR:
MID:
CFG:
PT:0
S:andy.sharp@lsi.com
RQ:
SSV:mhbs.lsil.com
NSV:
SSH:
R:<Larry.Scheer@lsi.com>,<Jan.Seidel@lsi.com>,<David.Olien@lsi.com>
MAID:2
X-Sylpheed-Privacy-System:
X-Sylpheed-Sign:0
SCF:#mh/Mailbox/sent
RMID:#imap/LSI/INBOX	0	DEC609CD0E54B2448DAF023C89AE9755EB50C3B1@cosmail02.lsi.com
X-Sylpheed-End-Special-Headers: 1
Date: Tue, 19 Jan 2010 14:09:10 -0800
From: Andrew Sharp <andy.sharp@lsi.com>
To: "Scheer, Larry" <Larry.Scheer@lsi.com>
Cc: "Seidel, Jan" <Jan.Seidel@lsi.com>, "Olien, David" <David.Olien@lsi.com>
Subject: Re: creating volumes
Message-ID: <20100119140910.56e67a32@ripper.onstor.net>
In-Reply-To: <DEC609CD0E54B2448DAF023C89AE9755EB50C3B1@cosmail02.lsi.com>
References: <6C678488C5CEE74F813A4D1948FD2DC7B7690B80@cosmail02.lsi.com>
	<D41A5B864986B546A595D85C4701A9C0B6E2DB15@cosmail03.lsi.com>
	<6C678488C5CEE74F813A4D1948FD2DC7B7690B96@cosmail02.lsi.com>
	<D41A5B864986B546A595D85C4701A9C0B6E2DB21@cosmail03.lsi.com>
	<6C678488C5CEE74F813A4D1948FD2DC7B7690BA9@cosmail02.lsi.com>
	<D41A5B864986B546A595D85C4701A9C0B6E2DB2C@cosmail03.lsi.com>
	<DEC609CD0E54B2448DAF023C89AE9755EB50C3B1@cosmail02.lsi.com>
Organization: LSI
X-Mailer: Sylpheed-Claws 2.6.0 (GTK+ 2.8.20; x86_64-pc-linux-gnu)
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Well, actually the standard way to do this is

reboot -s

from nfxsh.

I think maybe you've been messing in the PROM too much Larry ~:^)

Cheers,

a


On Mon, 18 Jan 2010 19:26:53 -0700 "Scheer, Larry"
<Larry.Scheer@lsi.com> wrote:

> If you answer no to both prompts to copy configuration information
> then when you boot from the newly created flash and start nfxsh for
> the first time you will be placed in the fist time install (FTI)
> routine where you are guided (using this term loosely here...)
> through the configuration of the filer.
>=20
> My bet is you are not booted from the flash that was just initialized
> and are booted from the old primary flash. To get to the newly
> initialized flash you need to stop the SSC boot sequence with a ctrl
> e to get into the prom and run:
>=20
> env view
> look at boot dev for this example lets say it is sda1
> next set boot_dev to the other flash:
>=20
> env set boot_dev sdb1
> reboot
>=20
> Then you will boot from the flash you just initialized.
>=20
> Larry
> ________________________________________
> From: Seidel, Jan
> Sent: Monday, January 18, 2010 5:43 PM
> To: Olien, David
> Cc: Scheer, Larry; Sharp, Andy
> Subject: RE: creating volumes
>=20
> Hi Dave,
>=20
> That=E2=80=99s a good question. I=E2=80=99m not really sure about it, so =
I=E2=80=99m copying
> Larry and Andy. I hope they can help.
>=20
> Short form: Dave did a flash_install and selected to not copy any
> configuration. However, after the flash_install his mgmt vsvr showed
> 2 volumes owned (which weren=E2=80=99t shown in the vol show output). Is =
this
> a bug?
>=20
> Thanks,
> Jan
>=20
> From: Olien, David
> Sent: Monday, January 18, 2010 5:33 PM
> To: Seidel, Jan
> Subject: RE: creating volumes
>=20
> Jan,
>=20
> When I ran flash_install.sh, I answered no to both of its questions
> about copying configuration.  So I would have expected it to not copy
> ANY configuration information from the previous flash ram.  Am I
> interpreting this incorrectly?
>=20
> Dave
>=20
> From: Seidel, Jan
> Sent: Monday, January 18, 2010 5:18 PM
> To: Olien, David
> Subject: RE: creating volumes
>=20
> Hi David,
>=20
> I=E2=80=99m not sure if it=E2=80=99s a flaw in the flash_install.sh. It l=
ooks like
> you already had this inconsistency between the =E2=80=9Cvsvr show=E2=80=
=9D and
> =E2=80=9Cvolume show=E2=80=9D output before the flash_install, is that co=
rrect? It
> might be that flash_install doesn=E2=80=99t fix these inconsistencies. Th=
ere
> might be an underlying defect somewhere, but therefore we would need
> to know what exactly was done with this system. It might be that it
> was just connected to different storage.
>=20
> Regards,
> Jan
>=20
>=20
> From: Olien, David
> Sent: Monday, January 18, 2010 5:04 PM
> To: Seidel, Jan
> Subject: RE: creating volumes
>=20
> Thanks You, Jan
>=20
> This worked.  I=E2=80=99ve created management and core volumes.
> Reading your email a little more closely.  The state I was in before,
> apparently the configuration was in a state where it wasn=E2=80=99t
> internally consistent?  The management vsvr thought it had management
> and core volumes, as displayed by the vsvr show command,  But those
> volumes didn=E2=80=99t appear in the volume show command output?
>=20
> So this would be a flaw in the install_flash.sh behavior.
>=20
> Dave
>=20
> From: Seidel, Jan
> Sent: Monday, January 18, 2010 4:48 PM
> To: Olien, David
> Subject: RE: creating volumes
>=20
> Hi David,
>=20
> Feel free to ask questions, I know that it=E2=80=99s not that easy to get
> familiar with everything.
>=20
> I think you were doing all the steps correctly. The mgmt vsvr is
> created automatically and the digits at the end are just a unique
> value for each filer (i.e. no 2 filers have the same mgmt vsvr name).
>=20
> The problem with your filer was that the mgmt vsvr still had the
> information about its mgmt and core volume. The flash_install must
> have somehow preserved that information. I  now deleted those 2
> volumes for you with the =E2=80=9Cvsvr delete force volume VOLNAME=E2=80=
=9D command.
> I copied the commands I ran below. This command can be used to
> recover from states like these (but it=E2=80=99s not the regular command =
to
> delete a volume, this is =E2=80=9Cvolume delete VOLNAME=E2=80=9D). It=E2=
=80=99s recommended
> to reboot the filer after the =E2=80=9Cvsvr delete force volume=E2=80=9D =
command.
> After the reboot you should then be able to create new volumes.
>=20
> If that doesn=E2=80=99t work or you run into any other problems please le=
t me
> know.
>=20
>=20
> g12r208 diag> vsvr show vs_mgmt_67462
> Virtual Server
> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
> ID                : 1
> NAS Gateway name  : g12r208
> Status            : Enabled
> Name              : VS_MGMT_67462
> NetBIOS name      : VS_MGMT_67462
> Pri. WinSrvr IP   : 0.0.0.0
> Sec. WinSrvr IP   : 0.0.0.0
> IP address 1      : 10.3.208.240/255.255.0.0
> Windows domain    : --Not-Set--
> Unix domain       : onstorlab
> LDAP domain       : --Not-Set--
> AutoCreate Path   : --Not-Set--
> AutoCreate VolID  : --Not-Set--
> Volumes owned     : vol_mgmt_67462 vol_core_67462
> Volumes mounted   : vol_mgmt_67462
> Read responses fragmentation  : default
> UDP checksum enabled          : yes
> Maximum mpx count             : default
> Strict name checking          : Enabled
>=20
>=20
> g12r208 diag> vsvr set VS_MGMT_67462
> g12r208 VS_MGMT_67462 diag> vsvr delete force volume vol_mgmt_67462
> Are you sure ? [y|n] : y
> Please reset the nas gateway after the command completes.
> Cluster DB records deleted
> g12r208 VS_MGMT_67462 diag> vsvr delete force volume vol_core_67462
> Are you sure ? [y|n] : y
> Please reset the nas gateway after the command completes.
> Cluster DB records deleted
> g12r208 VS_MGMT_67462 diag>
>=20
>=20
>=20
> Regards,
> Jan
>=20
> From: Olien, David
> Sent: Monday, January 18, 2010 4:33 PM
> To: Seidel, Jan
> Subject: creating volumes
>=20
> Sorry, Jan, I=E2=80=99m asking so many questions that seem obvious.
>=20
> I=E2=80=99m trying to create volumes on my cougar.
> I=E2=80=99ve installed new flash rams from the last nightly build,
> And rebooted using the nfxsh system reboot =E2=80=93s command.
>=20
> When I did the flash_install.sh, I answered NO to BOTH questions about
> Bringing configuration information forward from the existing flash.
> So  I would expect this machine to have NO configuration present.
>=20
> The first thing I notice is that there IS a =E2=80=9Cmanagement vsrvr=E2=
=80=9D
>=20
> G12r208 diag> vsvr show
> Virtual servers on nas gateway g12r208
>=20
>  ID     State                                                   Name
>=20
> 1         Enabled
> VS_MGMT_67462
>=20
> I am unable to disable or delete this vsvr.  I assume this is because
> this is the management vsvr for this node, And that it exists
> automatically after install (where does its name come from?) So I set
> my vsvr context to this vsvr. My prompt becomes
>=20
> G12r208 VS_MGMT_67462 diag>
>=20
> Doing a =E2=80=9Cvolume show=E2=80=9D returns nothing.
>=20
> Lun show all
>=20
> shows I have 6 LUNs available.  Two of them are named:
> ENGENIO_E3994A_R59_g12r208_mgmt
> ENGENIO_E3994A_R59_g12r208_core
>=20
> Which are both of Unknown Raid type, are in state =E2=80=9Cforeign_free=
=E2=80=9D, and
> Cluster N/A, and the Volume Column of the table is empty.
>=20
> So I =E2=80=9Clabel=E2=80=9D these Luns, to be of type Raid-0.
>=20
> They now show up as having Raid-0, in the free state, and belonging
> to Cluster q12r208.  Still no volume.
>=20
> Next I try to create the management volume.
> The =E2=80=9Cvolume show=E2=80=9D command still returns nothing.
> I try the command:
>=20
> System create mgmtVolume ENGENIO_E3994A_R59_g12r208_mgmt
>=20
> This fails with the message
>=20
> Operation failed. Volume already exists.
>=20
> What am I doing wrong here?
>=20
> Thanks!
> Dave
