How does one debug an Virtual Connect-FC flogin failure?

[chuckk281]

Thomas was looking to debug a SAN issue:

************************

I posted the note below regarding an issue with some blades in a VC-FC equipped enclosure not logging into the SAN fabric (get a “NOT-LOGGED-IN” message on HBA port status in VC screen- specifically three blades, one BL460G7 and two BL460G6, not able to log in on the bay 3 VC-FC module but fine on the bay 4 VC-FC module. Other blades in the same enclosure log in on both sides and this customers have 30+ additional enclosures connected to same SAN fabric with no issues.

We have a case opened with HP Support on this (4625176879) and they have had the customer upgrade all firmware, replace the HBA in some of the suspect blades, reset and reseat VC-FC modules, change slots where the suspect blades are inserted, etc. and we still have the NO-LOGGED-IN status. Right now support is requesting the log files from the Cisco IOS switches the VC-FC modules are connected to – but their SAN team says they never see a flogin request from the WWIDs for these servers.

So that leads me to ask – exactly how does one debug these flogin issues? I would have thought we could find a log entry somewhere in VC-FC module telling us why it would not log in, but we are finding nothing of interest. Can someone tell me some good steps to go about debugging this.

*************************

Chris was looking to help:

**********************

Has anyone ruled out a faulty transceiver (on either side) or cable?  If the upstream switch port isn’t getting an FLOGI, then that usually means one of a few things:

1. NPIV isn’t enabled (and this is done on a per VSAN basis with Cisco, or per port for Brocade)
2. There is an NPIV bug within the FC switches OS (check the version of firmware on the FC switches they are running)
3. Physical Layer issue (L1)

VC FC should see that the upstream switch isn’t reporting NPIV status which the port tries to do a PLOGI attempt.  I would have the customers SAN team check to see if they have the VCFC’s WWN logged into a port.  If so, and no other WWN’s are logging in, then that means there is a NPIV issue.  If the VC FC’s physical port hasn’t logged in, then that usually means a physical issue.

**************************

ANy other help or suggesitons? Could be a nasty one.

[Wverbraecken]

Hello,

Did you find a solution to your problem?

I'm installing a new C7000 enclosure with 10 BL490 G6 blades and 1 BL460 G6 blades. The enclosure also has 2 VC Flex 10 ethernet modules and 2 VC Fiber Channel 4Gbit modules.

All blades have a SAN connection on 2 ports except for the BL460 G6, it only has one port that is logged in to the fabric. I tried already change the blade to another position, recreating the server profile, downgrading and reïnstalling drivers, replacing the motherboard and replacing the Mezannine car, swapped the VC modules but nothing seems to help.

Thanks,

Wesley