LVM and VxVM
cancel
Showing results for 
Search instead for 
Did you mean: 

persistent DSF and new MDS switches

 
SOLVED
Go to solution
Highlighted
GH1
Occasional Advisor

persistent DSF and new MDS switches


Hi All,
I am new to the persistent DSF model for path links and have a question. We have a BL870c running HPUX 11.31 connected to an HP EVA8400 for storage using older MDS9216 switches via 2 fabric connections (fabric A and fabric B). We are getting ready to migrate from the 9216 to a set of 9148’s moving one fabric/path at a time (starting with fabric B). In the past when I have done this, additional steps were required on HPUX 11.11. I had to run ioscan, make note of the new paths (/dev/dsk/cXtXdX), and then modify the volume groups accordingly by removing the old paths with vgreduce and adding the new paths with vgextend. How does this process work with persistent DSF disks?
4 REPLIES 4
Kenan Erdey
Honored Contributor
Solution

Re: persistent DSF and new MDS switches

Hi,

the persistent DSF maps to the LUN WorldWide Identifier (WWID), a resource on the LUN similar
to a serial number. The WWID is unique to a LUN and does not change for that LUN. If any of a LUN lunpaths
change, the DSF bound to the LUN doesnot change â hence, the DSF has agile addressing.

so it should not be a problem. you are safe with persistent dsf.

http://www.scribd.com/doc/40054087/Persistent-Dsf-Migration
Computers have lots of memory but no imagination
Michael Leu
Honored Contributor

Re: persistent DSF and new MDS switches

You can check your WWIDs with
ioscan -C disk -P wwid
these are the same numbers you see in CV EVA at the vDisk.

As Kenan said there are no additional steps needed with 11.31 when you change your switches.
Ismail Azad
Esteemed Contributor

Re: persistent DSF and new MDS switches

Hi,

Irrespective of the target port or switch or adapter, the persistent DSF is bound to the LUN and does not record the bus it traverses accross or in simpler terms has nothing to do with a LUN path. Persistent DSF's "persist" accross reboots and hence, any topology change on the local server should not hamper the identification of that hardware in terms of the persistent DSF.

The LUN is "authenticated" in HPUX 11iv3 with a WWID speaking in terms of the persistent DSF and hence, adding to michael you could check your wwid in this fashion as well....

scsimgr get_attr -a wwid -H <64000/0xfa00/0x2>

Regards
Ismail Azad
Read, read and read... Then read again until you read "between the lines".....
GH1
Occasional Advisor

Re: persistent DSF and new MDS switches

Thank you all for the responses