can not see lun's

it's not similar to your previous issue ..with duplicate flogi ?

nope. What does

"Logged in --- NO"
"ON Fabric --- YES"

mean in the symmask list logins command?

Message was edited by:
protoss

The HBA is not logged into the FA but it does exist within the fabric.

I dont understand what could be the issue other than Zoning.

I completely removed the zones and re did zoning, when I did this I issued the symmask list logins command and saw that HBA was logged into the fabric. I asked the sys admin to restart the server and now it again says that it is not logged into the FA.

I checked masking and mapping every thing thrice.

This is another port that is also not logged into the fabric on the same FA but the other HBA's are logged into the same FA with out any problem.

Where could the issue be??

Are you using VCM's? Do you have devices mapped to that HBA in the VCM database?

symmaskdb -sid 61 list database -dir 7c -p 0 -wwn 50014380*******


[ Database Type : Type6
Last updated at : 11:25:05 AM on Wed Sep 10,2008

Director Identification : FA-7C
Director Port : 0

User-generated
Identifier Type Node Name Port Name Devices
---------------- ----- --------------------------------- ---------
5001438000****** Fibre 5001438000****** 5001438000****** 0265
0274
0283
0292
02A1
02B0
02BF
02CE
02DD
02EC
02FB
030A
0319
0328
0337
049F ]

Hope this explains.

I would like to add that in a few other FA ports also I see the same thing. The HBA is logged into the fabric and but not the FA. There is no consistancy at all.

for example in port 7A 1 HBA shows not logged into the FA, port 7C 2 HBA's not logged into the FA and list goes on.

Please let me know.

Message was edited by:
protoss

the ones that are showing this flaky behavior ...are they all VMware boxes or other platforms as well ?

Is the host HBA WWN is correct? What sort of HBA are you using?

The reason for asking is that 5001438000******

is a Hewlett Packard WWN. Given this is a VMWare (ie Intel based) box I would usually expect to see something like an Emulex or QLogic HBA.

M

HP re-branded Qlogic cards start with 50****. I have a bunch of those in my HP blade centers.

Fair enough! Though it's still worth checking the WWN is actually the correct one.

Also check the HBA driver & firmware is EMC qualified (Support Matrix) and that it's correctly configured (ie topology, link down timeout, etc).

M

Message was edited by:
MarcT

All of them are VMware boxes. Nothing else is connected to the DMX(It is a new one we bought) other than symcli host and it works fine.

Message was edited by:
protoss

Yes they are all Qlogic cards.

null

it's strange indeed. Can you think of anything else that these affected systems share? Same patch panel, same port group on the switch ?

HBA driver & firmware cant be the issue as all the VMware hosts have been built using the same hardware and s/w installations.

Most of them are working fine. should be something else.

This one is getting on to me.

Could add if there is more than one switch in the fabric, same ISL?