Our local hardware folks tell us the proper firmware is set for our two
FCP CHPID's.  Since we need to be connected thru a SAN switch (assuming
that requirement is still needed) the z800 Hardware Support folks
told us to try setting a SWITCH=   option on the CHPID defn in the
IOCP.  We tried that (even though none of the examples  show this)
and  got an IOCP error indicating that SWITCH= is not an option
on a CHPID type of FCP.   We're  cross posting this to the LINUX-390
list to see if there are any other suggestions before we go up the
HW support ladder.

Ron Greve
Admin and Research Computing
SDSU


>I have not done this personally, but your IOCP looks good to me provided
>your PARTITION and PATHs exist.  I don't believe point to point with
>TYPE=3DFCP is supported on a z/800, but you appear to be okay there.  I
>believe certain firmware is needed to support TYPE=3DFCP.  Are you sure
>your z/800 has that firmware?
>
>Regards,
>
>Kevin
>
>-----Original Message-----
>From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED] On
>Behalf Of Ron Greve
>Sent: Tuesday, November 07, 2006 12:39 PM
>To: [email protected]
>Subject: iocp /fcp question
>
>We are trying to get an FCP link set up from a   Z/800  IFL to an ESS
>800 via a QLOGIC SANBOX 5600 switch.
>
>The IOCP definition is as follows:
>
> CHPID PATH=3D12,PARTITION=3D(IFL),TYPE=3DFCP
> CHPID PATH=3D13,PARTITION=3D(IFL),TYPE=3DFCP
> CNTLUNIT CUNUMBR=3D1200,PATH=3D(12),UNIT=3DFCP
> CNTLUNIT CUNUMBR=3D1300,PATH=3D(13),UNIT=3DFCP
> IODEVICE ADDRESS=3D(1200,64),CUNUMBR=3D(1200),UNIT=3DFCP
> IODEVICE ADDRESS=3D(1300,64),CUNUMBR=3D(1300),UNIT=3DFCP
>
>When we activate the CHPID on the service console the CHPIDs indicate
>ONLINE, DEDICATED, NOT ISOLATED, DEFINITION ERROR.  The details screen
>for the CHPID indicate it is a Fibre Channel Converted CHPID Type.
>
>The hardware folks say the definition error indicates a mismatch
>between the Chpid Type and the IOCP definition.
>
>The ESS 800 connects to the switch without any problems.
>
>Is there something missing in our IOCP definition or is  the
>CHPID misconfigured in some way?
>
>Thanks for any suggestions.
>Ron Greve
>SDSU Admin and Research Computing

Reply via email to