Connecting new lpar to DS4800 storage server 1

[J1gh2]

Hi

I am trying to connect a new lpar to our DS4800 server but AIX is not seeing the logical disks. Here are the steps that I have taken so far:

1) I have configured logical drives on the DS4800 server and and added the host to it.
2) I have checked to ensure that the drivers are installed on aix. (there are fcs* and dac* are installed)

3) lslpp -l | grep fc* shows that I have the same drivers as on other boxes.
4) Connectivity to the DS4800 server is through fibre so I have checked the fibre cabling from the lpar into the switch which connects to the DS4800 and all looks ok.
5) however, I run cfgmgr but still can`t see the hdisks.

Is there anything else that I should be checking as well?

Many thanks for your help.

 

[mbourne2006]

What sort of zoning do you have in place?

 

[J1gh2]

Hi

It appears that there is port zoning in place. I have only recently taken over the SAN and the switches are not accessible remotely so I have not logged much on them. However, looking at the a screen dump here I see a two zone names with switch ports listed underneath so I assume it port zoning. What is the command to confirm the zoning from the comand line?

Thanks for your time

 

[mbourne2006]

If your switches are running Brocade FabricOS then try "zonehelp". The command you want is either "configshow" or "cfgshow". One of them is a generic query, the other is specific to zoning.

You can also use "nsshow" to print the nameserver table and see whether or not your LPAR HBA WWN/WWPN show up.

 

[J1gh2]

hi

Thank you very much. These commands are very useful. The line below is from switchshow, indicating that the WWN is seen on the switch and the port is online.

7 7 id N2 Online F-Port 10:00:00:00:c9:4e:60:e2

However, zoneshow indicates that port 7 is not included in the active zone. Below is the output from zoneshow:

IBM_2005_B32:admin> zoneshow
Defined configuration:
cfg: Config220805
DS4800_LPARS; UltriumTape_A
zone: DS4800_LPARS
1,0; 1,4; 1,1; 1,2; 1,3; 1,12; 1,13; 1,5
zone: UltriumTape_A
1,8; 1,9; 1,10
Effective configuration:
cfg: Config220805
zone: DS4800_LPARS
1,0
1,4
1,1
1,2
1,3
1,12
1,13
1,5
zone: UltriumTape_A
1,8
1,9
1,10

I would like to backup the config before trying to add the port to the zone but that would reboot the switch so I will plan downtime first and see how it goes.

Thanks a lot for your help.

 

[J1gh2]

Hi

I had a bit of a hairy morning with this. Since this is a production environment, I didn`t want to start playing with the active configuration (especially that I have not done this before) so I created a new zone and config and enabled it. I then found that I had lost all paths on that particular switch!!. Fortunately the LUNs failed over to the second switch in DS4800 so I have put them back to their preferred controllers having re-enabled the original config. Phew. My new lpar still cannot see the SAN though. I will plan for down time to get this done.

Is it risky to try adding the missing port to the current zone and config?

 

[mbourne2006]

Yikes - getting outside my rather limited knowledge now, I'm afraid. One comment I would make - all devices in a single zone will be "aware" of traffic generated by other devices in the same zone. So, if you develop a fault somewhere and something starts generating a lot of "registered state change notifications" then you can introduce problems on the other nodes in the same zone. We tend to try to restrict a zone to comprise: 1 HBA + 1 storage port. Downside - large configs ...

Question: did your new zone contain all ports in original zone + port #7 ?

Kind Regards,
Matthew Bourne
"Find a job you love and never do a day's work in your life.

 

[J1gh2]

Hi Matt

No, my new zone has only port 7 in it. I now intend to simply add my new zone to the effective configuration instead of adding port 7 to the original zone. You are right that "all devices in a single zone will be "aware" of traffic generated by other devices in the same zone". This happend to me yesterday and indeed introduced "problems on the other nodes in the same zone" when I made the changes.

I would like to make our configuration more resilient to changes so I may yet adopt your approach. Thanks for the tip.

Best regards

 

[J1gh2]

I have just realised that my lpar can only see through one path to the storage server.

fget_config –A shows:

---dar0---

dac0 ACTIVE dacNONE ACTIVE
hdisk2 dacNONE
hdisk3 dac0
hdisk4 dacNONE
hdisk5 dac0
hdisk6 dacNONE
hdisk7 dac0
hdisk8 dacNONE
hdisk9 dac0

I have run rmdev –Rdl dar0, fcs0, fcs1, dac0 and checked the host port information on the storage server is the same network address returned by lscfg –vl fc0 and fcs1. I then run cfgmgr again but I still get the above.

This is the sixth lpar on the frame and lslpp -l | grep devices.fcp* shows that the firmware level is the same as the other lpars (running aix 5.3) but I do not have this problem with any other lpar.

Any ideas please?

Thanks