Yes we currently do not have NPIV enabled (in the switches or CECs).  The way 
we have it setup for a cluster is to have the primary on one CEC and the backup 
on the other CEC.  That’s not an issue (the displays below the servers on 
different CECs, sorry).  I got the server mentioned to work by changing wwpn 
for the HDS USPv box on the adapters.  I flipped the wwpns.  No other servers 
are accessing these LUNs on this CEC.  Would setting up NPIV fix this issue?  
If so, can you point me to a good resource on setting up NPIV on zLinux?

Was
-----------
0.0.dc00        0x50060e800571f006      0x0017000000000000
0.0.dc00        0x50060e800571f006      0x0019000000000000
0.0.dd00        0x50060e800571f016      0x0017000000000000
0.0.de00        0x50060e800571f017      0x0017000000000000
0.0.df00        0x50060e800571f007      0x0017000000000000

Now
----------

0.0.dc00        0x50060e800571f007      0x0017000000000000
0.0.dc00        0x50060e800571f007      0x0019000000000000
0.0.dd00        0x50060e800571f017      0x0017000000000000
0.0.de00        0x50060e800571f016      0x0017000000000000
0.0.df00        0x50060e800571f006      0x0017000000000000

 


Thanks
Scott

-----Original Message-----
From: Linux on 390 Port [mailto:[email protected]] On Behalf Of òåôø áøåê
Sent: Tuesday, December 06, 2011 9:38 PM
To: [email protected]
Subject: Re: zFCP question for RHEL54

Hi,
i think that is exactly your problem.
If i got it right you are not using npiv. That means that all guests are using 
the same wwpn to the fabric on a spesific cec. 2 linux guests on the same cec 
cant access the same luns at the same time. The first guesr in the cec to 
access the luns wins. Either make sure you only share disks on different cecs 
and that no other guest is trying to access them, or enable npiv.
We enabled npiv causing each linux guest to have its own virtual wwpn. 
This make life simpler from clustering point of view. 

Hope this helps...
Offer Baruch 

Sent by emoze push mail


-----------------------------
From:   "Shumate, Scott" <[email protected]>
Subject:        Re: zFCP question for RHEL54
Date:           07 דצמבר 2011 00:14

There are 2 CECs involved. Each CEC has 4 fibre connections, 2 in one switch 
and 2 in another.  In our HDS box, we have 4 ports fibre cards (2 in one switch 
and 2 in the other).  The way its zoned is 4 connections from the CECs (2 from 
one CEC and 2 from the other) is zoned to 2 port on the HDS box.

So here is a break down

ZONE IN SWITCH1
CEC1 CHP 9C      
CEC1 CHP 9D
CEC2 CHP 48
CEC2 CHP 49
HDS PORT CL1-G
HDS PORT CL2-G

ZONE IN SWITCH2
CEC1 CHP 9E
CEC1 CHP 9F
CEC2 CHP 44
CEC2 CHP 47
HDS PORT CL1-H
HDS PORT CL2-H

To give you a break down, this server lives on CEC2 0.0.dc00 is attached to CHP 
44 0.0.dd00 is attached to CHP 47 0.0.de00 is attached to CHP 48 0.0.df00 is 
attached to CHP 49

I have another server that runs with the same zfcp.conf file.  It attaches with 
no issues.  That is what's confusing me.  Is it a zoning issue?


Thanks
Scott

-----Original Message-----
From: Linux on 390 Port [mailto:[email protected]] On Behalf Of Raymond 
Higgs
Sent: Tuesday, December 06, 2011 4:11 PM
To: [email protected]
Subject: Re: zFCP question for RHEL54

Scott,

> Dec  6 14:28:15 wil-zstsccdbtstbu01 kernel: zfcp: warning: failed 
> gid_pn nameserver request for wwpn 0x50060e800571f006 for adapter
0.0.dc00
> Dec  6 14:28:15 wil-zstsccdbtstbu01 kernel: zfcp: port erp failed 
> (adapter 0.0.dc00, wwpn=0x50060e800571f006)

The nameserver is a service that runs on your switches.  GID_PN is a request to 
the nameserver for Get Port Identifier by Port Name.  The error is that switch 
doesn't know about target port 0x50060e800571f006 so it can't return the n-port 
id that Linux needs.

I'd investigate like this...

1. check zoning on your switch.  Your channel may not have permission to talk 
to that target port.
2. check NPIV settings.  If NPIV is on and you meant for it to be off, then 
that may change how you zone at the switch.
3. check cabling.  If these linux systems are running on 2 different CECs, then 
maybe some channels on 1 CEC aren't cabled the way you expect.

Regards,

Ray Higgs
System z FCP Firmware Development
Bld. 706, B42
2455 South Road
Poughkeepsie, NY 12601
(845) 435-8666,  T/L 295-8666
[email protected]

Linux on 390 Port <[email protected]> wrote on 12/06/2011 03:04:53
PM:

> From: "Shumate, Scott" <[email protected]>
> To: [email protected]
> Date: 12/06/2011 03:11 PM
> Subject: zFCP question for RHEL54
> Sent by: Linux on 390 Port <[email protected]>
>
> I'm running into issues adding zFCP LUNS to a clustered zLinux server.

> The primary servers sees the disk with no issues and have 
> multipath'ing turned on.  Both servers should be pointing to the same 
> scsi disk.  What am I missing?  Any ideas?  I've tried rebooting with 
> no results.  I do see the following error in
/var/log/messages
>
> Dec  6 14:28:15 wil-zstsccdbtstbu01 kernel: zfcp: warning: failed 
> gid_pn nameserver request for wwpn 0x50060e800571f006 for adapter
0.0.dc00
> Dec  6 14:28:15 wil-zstsccdbtstbu01 kernel: zfcp: port erp failed 
> (adapter 0.0.dc00, wwpn=0x50060e800571f006)
>
> I see this for each adapter.  I did try to remove the unit and port 
> vary the adapters offline and tried it again with the same results.
>
> The server that works
>
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>
> [root@wil-zstsccdbtst01 ~]# lsscsi
> [0:0:0:1]    disk    HITACHI  OPEN-V           6008  /dev/sda
> [0:0:0:2]    disk    HITACHI  OPEN-V           6008  /dev/sde
> [1:0:0:1]    disk    HITACHI  OPEN-V           6008  /dev/sdb
> [2:0:0:1]    disk    HITACHI  OPEN-V           6008  /dev/sdc
> [3:0:0:1]    disk    HITACHI  OPEN-V           6008  /dev/sdd
>
> [root@wil-zstsccdbtst01 ~]# lszfcp -D
> 0.0.dc00/0x50060e800571f006/0x0017000000000000 0:0:0:1 
> 0.0.dc00/0x50060e800571f006/0x0019000000000000 0:0:0:2 
> 0.0.dd00/0x50060e800571f016/0x0017000000000000 1:0:0:1 
> 0.0.de00/0x50060e800571f007/0x0017000000000000 2:0:0:1 
> 0.0.df00/0x50060e800571f017/0x0017000000000000 3:0:0:1
>
> [root@wil-zstsccdbtst01 ~]# cat /etc/zfcp.conf
> 0.0.dc00        0x50060e800571f006      0x0017000000000000
> 0.0.dd00        0x50060e800571f016      0x0017000000000000
> 0.0.de00        0x50060e800571f007      0x0017000000000000
> 0.0.df00        0x50060e800571f017      0x0017000000000000
> 0.0.dc00        0x50060e800571f006      0x0019000000000000
>
>
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>
> The server that fails
>
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> [root@wil-zstsccdbtstbu01 0x50060e800571f006]# lsscsi
> [root@wil-zstsccdbtstbu01 0x50060e800571f006]#
>
> [root@wil-zstsccdbtstbu01 0x50060e800571f006]# lszfcp -D
> Error: No fcp devices found.
>
> [root@wil-zstsccdbtstbu01 0x50060e800571f006]# cat /etc/zfcp.conf
> 0.0.dc00        0x50060e800571f006      0x0017000000000000
> 0.0.dd00        0x50060e800571f016      0x0017000000000000
> 0.0.de00        0x50060e800571f007      0x0017000000000000
> 0.0.df00        0x50060e800571f017      0x0017000000000000
> 0.0.dc00        0x50060e800571f006      0x0019000000000000
>
>
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>
>
>
> Thanks
> Scott
>
> -----Original Message-----
> From: Linux on 390 Port [mailto:[email protected]] On Behalf Of 
> John Campbell
> Sent: Monday, December 05, 2011 9:25 AM
> To: [email protected]
> Subject: At the optometrist's ...
>
> http://www.datamation.com/imagesvr_ce/5678/geek-eye-doctor.jpg  ;-)
>
> My first though was whether there'd be enough humor in an EBCDIC
version.
>
> -soup
>
> --
> John R. Campbell         Speaker to Machines          souperb at
> gmail dot com MacOS X proved it was easier to make Unix user- friendly

> than to fix Windows
>
> ----------------------------------------------------------------------
> For LINUX-390 subscribe / signoff / archive access instructions, send 
> email to [email protected] with the message: INFO LINUX-390 or 
> visit http://www.marist.edu/htbin/wlvindex?LINUX-390
> ----------------------------------------------------------------------
> For more information on Linux on System z, visit
http://wiki.linuxvm.org/
>
> ----------------------------------------------------------------------
> For LINUX-390 subscribe / signoff / archive access instructions, send 
> email to [email protected] with the message: INFO LINUX-390 or
visit
> http://www.marist.edu/htbin/wlvindex?LINUX-390
> ----------------------------------------------------------------------
> For more information on Linux on System z, visit 
> http://wiki.linuxvm.org/
>

----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions, send email to 
[email protected] with the message: INFO LINUX-390 or visit 
http://www.marist.edu/htbin/wlvindex?LINUX-390
----------------------------------------------------------------------
For more information on Linux on System z, visit http://wiki.linuxvm.org/

----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions, send email to 
[email protected] with the message: INFO LINUX-390 or visit 
http://www.marist.edu/htbin/wlvindex?LINUX-390
----------------------------------------------------------------------
For more information on Linux on System z, visit http://wiki.linuxvm.org/


----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions, send email to 
[email protected] with the message: INFO LINUX-390 or visit 
http://www.marist.edu/htbin/wlvindex?LINUX-390
----------------------------------------------------------------------
For more information on Linux on System z, visit http://wiki.linuxvm.org/

Reply via email to