From: Chris Ruhnke (ruhnke@us.ibm.com)
Date: Tue Dec 16 2003 - 15:51:46 EST
OK all you hardware geeks, limber up your little grey cells...
My Tru64 configuration includes two pairs of shared HSZ50 controllers.
At boot time one pair identifies its "disks" [actually RAID-5 sets] as:
pza2 at pci1 slot 4
pza2 firmware version: DEC P01 A12
scsi6 at pza2 slot 0
rzb49 at scsi6 target 1 lun 1 (LID=3D59) (DEC HSZ50-AX V51Z)(Wide16)
rzc49 at scsi6 target 1 lun 2 (LID=3D60) (DEC HSZ50-AX V51Z)(Wide16)
rzd49 at scsi6 target 1 lun 3 (LID=3D61) (DEC HSZ50-AX V51Z)(Wide16)
rzb50 at scsi6 target 2 lun 1 (LID=3D67) (DEC HSZ50-AX V51Z)(Wide16)
rzc50 at scsi6 target 2 lun 2 (LID=3D68) (DEC HSZ50-AX V51Z)(Wide16)
Note the "DEC HSZ50-AX V51Z" ident string.
The other pair identifies its "disks" as:
pza3 at pci1 slot 5
pza3 firmware version: DEC P01 A12
scsi7 at pza3 slot 0
rz56 at scsi7 target 0 lun 0 (LID=3D82) (... ) (Wide16)
rz57 at scsi7 target 1 lun 0 (LID=3D90) (... ) (Wide16)
rz58 at scsi7 target 2 lun 0 (LID=3D98) (... ) (Wide16)
Note the "..." ident string. Actually it seems to be unprintable binary
codes -- one of which must be an X-OFF character because after displaying
this text, the console terminal input is "frozen" and I have to go into
setup mode and "reset communications" before I can type commands again.
The difficulty is that these servers (two AS8200's) are connected to a
console terminal concentrator via DS200 terminal servers which DO NOT
like to be X-OFFed.
I have looked everywhere I can think on the 2nd pair of HSZ50's and cannot
find where these ident strings are generated. The only thing I can find
is a difference in the firmware levels. A "SHOW THIS" on the 1st pair of
controllers identifies the controllers as "HSZ50-AX <serial number>
Firmware V51Z-0, Hardware A01". The 2nd HSZ50 pair has a different rev
level. "SHOW THIS" identifies the controllers as "HSZ50-AX <serial
number>
Firmware V57Z-0, Hardware C02".
Does anyone know how I can fix the 2nd pair of controllers to ident
correctly? If I can fix this problem, I can finally run this cluster
from a remote location (i.e. my basement, logged in to the console
concentrator and connected dynamically to the console ports) and not
have to drive in to the data center every time shutdown/reboots are
required.
--CHRis
Chris H. Ruhnke
Systems Management Integration Professional
Mid-Range Technical Services
IBM Global Services
St. Louis, MO
O'Toole's Law: Murphy is an optimist.
This archive was generated by hypermail 2.1.7 : Sat Apr 12 2008 - 10:49:46 EDT