ESXi 4.1u1 iSCSI install

[Crusnik01]

Hi!

I'm trying to install ESXi 4.1 Update 1 to a iSCSI LUN, which I will then boot my diskless server from.

I have managed to get the ESXi installation/setup to a iSCSI LUN, so that I might do a 'sanboot ...' after a 'sanhook'.

But when I sanboot my ESXi installation, it can't see any disks. If I look at my QNAP console (iSCSI target), the target I just hood has been released (not initiatet), even though it was previously.

How can I get the iSCSI connection to persist over to my sanboot?

Here are the commands I run:

Code:

vcreate --tag 100 net0
ifopen net0-100
dhcp net0-100
sanhook --drive 0x81 ${root-path} // Root-Path = iscsi:10.1.4.10:::0:iqn.2004-04.com.qnap:ts-859:iscsi.esx02.c0b5b2
sanboot --no-describe iscsi:10.1.4.10:::0:iqn.2004-04.com.qnap:ts-859:iscsi.esxi41u1.c0b5b2

And everything works fine, I get into the ESXi installation, but it tells me there are no supported devices to write the image to. And as I have mentioned, when I look in the QNAP control panel, I see that the iSCSI target has become unmounted.

Hope someone can help! =)

[EDIT]
Okey, it seems that my initial thought of the connection cutting when I do the sanboot isn't quite right.

It seems to be more "timeout" based, or something. If I do a 'sanhook', and just let the connection sit for a minute or two, it disconnects.

iPXE still thinks the drive is there, but it's not connected (checking the QNAP admin console). If I do a unhook, and the hook it up again, same thing.

I see that the iSCSI connection actually persists a good while into the ESXi setup initialization, but cuts at random times (I've done the test a few times).

Is this a value I can set in sanhook? Why does it timeout like that?

Log entries from the iSCSI target below:

Code:

    2011-06-11    20:33:07    iqn.2010-04.org.ipxe:esx02    10.1.2.2    ---    iSCSI    iqn.2004-04.com.qnap:ts-859:iscsi.esx02.c0b5b2    Logout
    2011-06-11    20:31:52    iqn.2010-04.org.ipxe:esx02    10.1.2.2    ---    iSCSI    iqn.2004-04.com.qnap:ts-859:iscsi.esx02.c0b5b2    Login OK
    2011-06-11    20:29:41    iqn.2010-04.org.ipxe:esx02    10.1.2.2    ---    iSCSI    iqn.2004-04.com.qnap:ts-859:iscsi.esx02.c0b5b2    Logout
    2011-06-11    20:28:26    iqn.2010-04.org.ipxe:esx02    10.1.2.2    ---    iSCSI    iqn.2004-04.com.qnap:ts-859:iscsi.esx02.c0b5b2    Login OK

This is clearly something programatic. Here is a New-TimeSpan result (PowerShell):

Code:

PS C:\> New-TimeSpan 20:31:52 20:33:07


Days              : 0
Hours             : 0
Minutes           : 1
Seconds           : 15
Milliseconds      : 0
Ticks             : 750000000
TotalDays         : 0,000868055555555556
TotalHours        : 0,0208333333333333
TotalMinutes      : 1,25
TotalSeconds      : 75
TotalMilliseconds : 75000



PS C:\> New-TimeSpan 20:28:26 20:29:41


Days              : 0
Hours             : 0
Minutes           : 1
Seconds           : 15
Milliseconds      : 0
Ticks             : 750000000
TotalDays         : 0,000868055555555556
TotalHours        : 0,0208333333333333
TotalMinutes      : 1,25
TotalSeconds      : 75
TotalMilliseconds : 75000

Both entries are exactly 75 ticks. The question now, is the issue iPXE based or is it the target (QNAP)?

I run several other iSCSI connections (another ESXi server, and windows hosts), no issues there.

[MultimediaMan]

Check out ESXi v4.1 Update 2 or newer... you will be pleasantly surprised.

Code:

#!ipxe
goto startme
:retry
sleep 5
:startme
set keep-san 1
set initiator-iqn iqn.2011-04.local.us.ar.bv.olympus:v-ipxe-a.olympus.bv.ar.us.local
sanboot --drive 0x80 iscsi:atlas.olympus.bv.ar.us.local::::iqn.1991-05.com.microsoft:atlas-v-ipxe-a-target ||
sanboot --drive 0x81 --no-describe http://epimetheus.olympus.bv.ar.us.local/vmware/4.1/x86_64/iso/ESXi_4_1_U2.iso || goto retry