VIO Issue
- From: "Copper, Steve" <scopper@xxxxxxxxxxxxxxxxxx>
- Date: Mon, 10 Apr 2006 13:30:31 +0100
Hi All,
I have a question about multipathing on the VIO servers which I hopeOne other thing that I have noticed is that when listing the filesets which
someone can help with. Basically I am trying to set up the scenario as per
the attached diagram which is from the redbook sg247940, Advanced Power
Virtualization on IBM System p5, Chapter 5, page 249.
The only difference to the diagram is that the DS4800 goes through 2 SAN
switches, one controller through each. Each fibre card on the VIO is zoned
to see only one of the controllers. I have created a host group on the
ds4800 which has the 2 fibre cards in and 12 disks assigned to the host
group. As per the diag ds4800 attached.
The issue I am having is that in this configuration, after I have done the
cfgdev on the VIO server, I am seeing 2 hdisks for the same disk (I have
checked the ieee_volname parm to be sure). In this case I have
hdisk0-hdisk28 inclusive. Hdisk0, hdisk1 and hdisk2 are the internal
disks. Not sure what hdisk3 and hdisk4 are doing here, they were present
even before I allocated and disks to the VIO. Then I have hdisk5-28 which
are the 24 paths to the 12 disks.
As per the diagram, it indicates that there should only be 1 hdisk for the
disk as the RDAC driver will have sorted out the dual path to present only
the one disk. This is the part that I am not seeing and am stuck on.
Earlier on in the chapter on page 245 there is a Note which says to use
the oem_setup_env command (which creates a root shell) for installing and
configuring the multi-path environment on the VIOs. What commands are used
in the above shell to achieve the configuration?
are required for the RDAC there is no version number or description
associated with them, as below. Is this right? Or should there be that info?
$ lssw | grep fc
devices.common.IBM.fc.hba-api
devices.common.IBM.fc.rte
devices.fcp.disk.array.diag
devices.fcp.disk.array.rte
devices.fcp.disk.rte 5.3.0.41 A F FC SCSI CD-ROM, Disk,
devices.fcp.tape.rte 5.3.0.41 A F FC SCSI Tape Device
Software
Ioslevel on the VIOs is 1.2.1.1-FP-7.2
Thanks in advance for your replies.
Regards
Steve Copper
<<mpio.JPG>> <<ds4800.JPG>>
______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
For more information please visit http://www.messagelabs.com/email
______________________________________________________________________
Attachment:
mpio.JPG
Description: JPEG image
Attachment:
ds4800.JPG
Description: JPEG image
- Follow-Ups:
- Re: VIO Issue
- From: John Monahan
- Re: VIO Issue
- Prev by Date: Re: VIO Issue
- Next by Date: Sajjad S Hashimi is out of the office until 4/11/06
- Previous by thread: Re: VIO Issue
- Next by thread: Re: VIO Issue
- Index(es):
Relevant Pages
- Re: VIO Issue
... Subject: VIO Issue ... I have a question about multipathing on the VIO servers
which I hope ... ds4800 which has the 2 fibre cards in and 12 disks assigned to the host
... The issue I am having is that in this configuration, ... (AIX-L) - MPXIO: Cant get to encapsulate drives ...
... I can't get MPXIO to encapsulate the following pairs of disks ... are derived
from a Clariion FC4700 unit (through two Brocade ... # Load balancing global configuration:
... # List of ULP modules for loading during port driver attach time ... (comp.unix.solaris) - Re: VIO Issue
... John Monahan ... VIO Issue ... ds4800 which has the 2 fibre cards
in and 12 disks assigned to the host ... This email has been scanned by the MessageLabs
Email Security System. ... (AIX-L) - Re: Rebuild Exchange 2003 SP1 on same hardware
... You can (depending on your disk configuration) also leave them. ... rebuilding
the server and assigning a drive letter back to the disks, ... Exchange Servers configuration
container information from AD. Meaning you ... Depending on what I highlighted in response
to 1) You can either restore ... (microsoft.public.exchange.admin) - md mirror / ext3 / dual core performance strange phenomenon?
... I have this configuration: ... starting from -j4 the compile time suddenly
goes to ... I tried changing the readahead value of all relevant devices and... ...
I kicked one of the two disks out of the mirror. ... (Linux-Kernel)