Hi  TSMers,

just some hints for those (like me) who are (were) hesisating between LP8000
and IBM 622x ?

Except for the price side, LP8000s come with their own non-IBM drivers...
I had some difficulties to make them work...

LP8000 and 622x are belongs to the same family.

LP6000/LP7000 is recognized under AIX as devices.pci.df1000f7 (f700)
LP8000 is recognized under AIX as devices.pci.df1000f8 (f800)  (Emulex
jumper set)
LP8000 is recognized under BULL AIX as devices.pci.df1001f8 (f801)  (bull
jumper set)
LP900x is recognized under AIX as devices.pci.df1000f9

IBM supports (they told me) the LP8000 but there is not the least driver
called .df1000f8
you can only find ...f7 and ...f9 ones.

Here is my experience (many thanx again to my STK support guy !)

remove all fc drivers (including IBM.common.fc)
install devices.pci.lpfc.* stuff

if you are in a point-to-point connection (LP8000 - fibre - T9840), leave
the ODM value 'topology'
set  to 4
if you are in a switched fabric (LP8000 - fibre - StorageNet 41xx -
fibre - T9840), change the ODM value 'topology'
set  to 2
reboot

manually create your TSM mtX device(s) with the mkdev command..

However, I still have TWO pbs :
---------------------------------
1) I can not see any drive through the SAN, my switch is well configured,
1) I can not see any drive through the SAN, my switch is well configured,
all my ports are F_ports.
I filled the /etc/lpfc.conf with the WWW of my drives associated to the
desired target by neither
a cfgmgr nor a reboot change anything.

Can someone send me his /etc/lpfc.conf

2) As direct conection uses a LOOP protocol, if  I change something on your
lpfcX driver of rmtX drive
(for instance, block_size), I'm not able to make the device available again
with a 'mkdev -l lpfcX'

I must reboot !

Thanx for your help
Herve (with Dimon)