Disk Enclosures
1748185 Members
4376 Online
108759 Solutions
New Discussion юеВ

XP256 and Fibre Channel configuration

 
SOLVED
Go to solution
SimonD
Advisor

XP256 and Fibre Channel configuration

Hello,
I have a L3000 system with a Tachyon Fibre Channel adapter connected to an XP256.
I configured 16 LUN on a fibre channel interface of the XP256.
with the XP's configuration software I assigned target ID 7
(the only possible with FC interfaces) to all LUNs and device ID from 00 to 0F.
On HP-UX side, I think to find 16 disks with the same controller, 2 different target ID:
c7t0d0 c7t0d1 c7t0d2 c7t0d3 c7t0d4 c7t0d5 c7t0d6 c7t0d7
c7t1d0 c7t1d1 c7t1d2 c7t1d3 c7t1d4 c7t1d5 c7t1d6 c7t1d7
(or something like this)

But on the UX machine, with the ioscan command I can see only the first 8 LUNs:
/dev/dsk/c7t0d0
/dev/dsk/c7t0d1
/dev/dsk/c7t0d2
/dev/dsk/c7t0d3
/dev/dsk/c7t0d4
/dev/dsk/c7t0d5
/dev/dsk/c7t0d6
/dev/dsk/c7t0d7
How I can access the next 8 LUNs?
Many thanks
Simon
12 REPLIES 12
Victor BERRIDGE
Honored Contributor

Re: XP256 and Fibre Channel configuration

Hi
What OS have you on your L3000, and what driver are you using (FC-AL?)
Is your disk bay a true XP256 or an HDS7700E?

Best regards
Victor
SimonD
Advisor

Re: XP256 and Fibre Channel configuration

Hi Victor,
you are right: the disk is a Hitachi 7700E, but is the first time I see it and I called it XP256.
The OS is HP-UX 11 and the FC adapter is a Tachyon PCI (driver version B.11.00.05).
Ciao
Simone
Victor BERRIDGE
Honored Contributor

Re: XP256 and Fibre Channel configuration

Hi Simone,
Im in almost the same situation with a HDS9960 (eq XP512).
It is a HP imposed limitation to give advantage to XPXXX with HP-UX environment on the original HDS7700-99XX, the limitation comes from the way the disks are seen by the HP HBA => hitachi-open=> 8 Lun PER PORT
=> XP no restriction...
As you have noticed, the diffrences is the possibility to increment t when d has reached 7...

2 workarounds:
-)Negociate with HP for XP upgrade (microcode)
-)Use a JNI Host Bus Adapter (HP doesnt like this at alll...), Yes it works but you will have no support from HP, so its not advisable if you have something sensible running with MC-ServiceGuard, but Im sure HDS will give all their support fromHBA too HDS770E.

HP USA seem to not see the difficulties they are giving to their Europeen collegues who have to deal with this situation, as it seems that in Europe the tendancy when you say Im "OPEN" is to have all the flavours possible on a same site, they forget that we had to choose between a IBM ESS and a HDS bay, we didnt realize and have not noticed this limitation at the time of deciding favoring HDS because we have plenty on little and medium HDS subsystem mostly plugged to HP, but also SUN and IBM aix. We couldnt choose an XP because HP doesnt (at the time)support Mainframes and we have ESCON port to an IBM mainframe.
We were thinking of acquiring 2 second N4000, but since I plugged 2 L2000, I am waiting a decision about my N4000 I cannot connect..., so now we are buying SUNs and IBMs instead, and a big project that might have been on a superdome may finish on a SUN E10000, and I will have to slowly get rid of HPs, a pity...

I wish you good luck as you will need a lot..
All the best
Victor

P.S.
I did try the JNI card (128LUNS per HBA), and yes it works but the direction is on the way to decide if we are to keep HP or not...
Bernd Reize
Trusted Contributor

Re: XP256 and Fibre Channel configuration

Hi Simone,
I'm not sure, if it is the same with a HDS, but with a XP there is a configuration option wich might solve your problem:
The XP allows several so called Host Modes, wich influence the way, FC to SCSI Mapping is done.
HP recommends mode 0x18 (HP UltraSCSI), this mode should do the right mapping with 7 LUNs per SCSI ID.

Greets, Bernd
SimonD
Advisor

Re: XP256 and Fibre Channel configuration

Hi Bernd
I haven't already tested the solutions that you provided, but I have an official answer from Hitachi: HP-UX doesn't support HDS7700E.
The firmware of the HDS7700E is different from the XP256 one, and the driver of the Tachyon FC adapter recognize it and blocks the LUNs over the eight. This is due for commercial purposes (HP want to sell XP with his HP-UX systems). The only solution is to change the Tachyon adapter with a JNI (as Victor suggested).
From HP side, I haven't already received a definitive response...
SimonD
Advisor

Re: XP256 and Fibre Channel configuration

Here is the HP Labs' answer. Do you know how I can change from PDA to VSA?!?

... After looking at the information
> that you provided, it appears that some of the interfaces in the 7700e are
> not properly configured. The following devices are configured for
> Peripheral Device Addressing (this is determined by the presence of the
> 255 in the HW address).
>
> Apr 26 18:28:56 hostname vmunix: 3/2/0.8.0.255.0.0.0 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/2/0.8.0.255.0.0.1 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/2/0.8.0.255.0.0.2 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/2/0.8.0.255.0.0.3 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/2/0.8.0.255.0.0.4 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/2/0.8.0.255.0.0.5 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/2/0.8.0.255.0.0.6 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/2/0.8.0.255.0.0.7 sdisk
>
> and
>
> Apr 26 18:28:56 hostname vmunix: 5/0/0.8.0.255.0.0.0 sdisk
> Apr 26 18:28:56 hostname vmunix: 5/0/0.8.0.255.0.0.1 sdisk
> Apr 26 18:28:56 hostname vmunix: 5/0/0.8.0.255.0.0.2 sdisk
> Apr 26 18:28:56 hostname vmunix: 5/0/0.8.0.255.0.0.3 sdisk
> Apr 26 18:28:56 hostname vmunix: 5/0/0.8.0.255.0.0.4 sdisk
> Apr 26 18:28:56 hostname vmunix: 5/0/0.8.0.255.0.0.5 sdisk
> Apr 26 18:28:56 hostname vmunix: 5/0/0.8.0.255.0.0.6 sdisk
> Apr 26 18:28:56 hostname vmunix: 5/0/0.8.0.255.0.0.7 sdisk
>
> Using Peripheral Device Addressing (PDA) only allows for 8 LUNs per SCSI
> channel (this is a limitation of hp-ux). To handle the large number of
> LUNs possible on the newer disk array products, a scheme called Volume Set
> Addressing (VSA) was devised. In this scheme, "virtual" SCSI busses are
> created per physical hardware path, thus allowing more than 8 LUNs per
> physical path.
>
> Apr 26 18:28:56 hostname vmunix: 3/1/0.3 tgt <-------Virtual SCSI bus
> Apr 26 18:28:56 hostname vmunix: 3/1/0.3.0 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.3.1 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.3.2 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.3.3 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.3.4 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.3.5 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.4 tgt <-------Virtual SCSI bus
> Apr 26 18:28:56 hostname vmunix: 3/1/0.4.0 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.4.1 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.4.2 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.4.3 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.4.4 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.4.5 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.5 tgt <-------Virtual SCSI bus
> Apr 26 18:28:56 hostname vmunix: 3/1/0.5.0 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.5.1 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.5.2 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.5.3 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.5.4 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.5.5 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.6 tgt <-------Virtual SCSI bus
> Apr 26 18:28:56 hostname vmunix: 3/1/0.6.0 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.6.1 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.6.2 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.6.3 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.6.4 sdisk
> Apr 26 18:28:56 hostname vmunix: 3/1/0.6.5 sdisk
>
> In the above output, we have one Fibre Channel adapter in slot 3/1/0 in
> the V-class. Configured on that adapter there are 4 virtual SCSI buses
> present. On each of those busses, we could have 6 LUNs configured ( it is
> possible to have up to 8 per virtual bus).
>
> In this case it appears that only the 7700e interfaces attached to 3/2/0
> and 5/0/0 are configured for PDA, the other interfaces appear to be set
> properly. The configuration of the disk device determines which addressing
> scheme will be used. If the array is configured for PDA, then you will
> only see 8 LUNs maximum on that hardware path.
Victor BERRIDGE
Honored Contributor
Solution

Re: XP256 and Fibre Channel configuration

Hi Simone,
Thanks for the info, I will search also...

Here is the output for the JNI I tested...

All the best
Victor

# ioscan -funC disk
Class I H/W Path Driver S/W State H/W Type Description
============================================================================
disk 2 0/0/1/0.1.0 sdisk CLAIMED DEVICE HP DVD-ROM 304
/dev/dsk/c0t1d0 /dev/rdsk/c0t1d0
disk 0 0/0/2/0.6.0 sdisk CLAIMED DEVICE SEAGATE ST318404LC
/dev/dsk/c1t6d0 /dev/rdsk/c1t6d0
disk 1 0/0/2/1.6.0 sdisk CLAIMED DEVICE SEAGATE ST318404LC
/dev/dsk/c2t6d0 /dev/rdsk/c2t6d0
disk 11 0/4/0/0/5/0.0.0 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t0d0 /dev/rdsk/c4t0d0
disk 12 0/4/0/0/5/0.0.1 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t0d1 /dev/rdsk/c4t0d1
disk 13 0/4/0/0/5/0.0.2 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t0d2 /dev/rdsk/c4t0d2
disk 14 0/4/0/0/5/0.0.3 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t0d3 /dev/rdsk/c4t0d3
disk 15 0/4/0/0/5/0.0.4 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t0d4 /dev/rdsk/c4t0d4
disk 16 0/4/0/0/5/0.0.5 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t0d5 /dev/rdsk/c4t0d5
disk 17 0/4/0/0/5/0.0.6 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t0d6 /dev/rdsk/c4t0d6
disk 18 0/4/0/0/5/0.0.7 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t0d7 /dev/rdsk/c4t0d7
disk 19 0/4/0/0/5/0.1.0 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t1d0 /dev/rdsk/c4t1d0
disk 20 0/4/0/0/5/0.1.1 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t1d1 /dev/rdsk/c4t1d1
disk 21 0/4/0/0/5/0.1.2 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t1d2 /dev/rdsk/c4t1d2
disk 22 0/4/0/0/5/0.1.3 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t1d3 /dev/rdsk/c4t1d3
disk 23 0/4/0/0/5/0.1.4 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t1d4 /dev/rdsk/c4t1d4
disk 24 0/4/0/0/5/0.1.5 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t1d5 /dev/rdsk/c4t1d5
disk 25 0/4/0/0/5/0.1.6 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t1d6 /dev/rdsk/c4t1d6
disk 26 0/4/0/0/5/0.1.7 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t1d7 /dev/rdsk/c4t1d7
disk 27 0/4/0/0/5/0.2.0 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t2d0 /dev/rdsk/c4t2d0
disk 28 0/4/0/0/5/0.2.1 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t2d1 /dev/rdsk/c4t2d1
disk 29 0/4/0/0/5/0.2.2 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t2d2 /dev/rdsk/c4t2d2
disk 30 0/4/0/0/5/0.2.3 sdisk CLAIMED DEVICE HITACHI OPEN-9
/dev/dsk/c4t2d3 /dev/rdsk/c4t2d3
#
SimonD
Advisor

Re: XP256 and Fibre Channel configuration

Hi Bernd,
I tested your suggestion, but don't works....
I tried also with 0x00 and 0x0F with the same result.
many thanks
Simone
SimonD
Advisor

Re: XP256 and Fibre Channel configuration

Hi Victor,
I'm waiting a JNI interface (an Hitachi gift, eheheh...) I hope it arrives soon: it seems the only solution.
In the same time I'm trying to decode HP Labs' answer....