Hi Guys

I got in to a very peculiar problem when I was testing FCOE initiator with 
Open-Fcoe target which is running in VN2VN configuration.

Connections are done back to back  :  Host Initiator  (ESXi) -> Target  (RHEL 
Linux LIO Open-FCoE Target, Intel x520 Niantic card)

When ESXi initiator is doing I/O to the LUN from Open-FCoE Target machine, 
initiator received inconsistent data (mismatch of SEQ_ID), which resulted in 
sending an abort for OXID : 11B7, sent an ABORT to the target.

Target Responded back with BA_ACC on altogether a different command with OXID: 
11cb.

We have never sent an ABORT for OXID: 11CB but abort was sent for a different 
IO, which caused the initiator to suspend activity.

Please see attached document which has Wireshark frames collected on RHEL Linux 
LIO Open-FCoE Target machine.

What I would like to ask is, does this behavior ever observed on open-fcoe 
target? if it is a known issue any patch available to patch the target machine?

Thanks in advance for your help and time in to this.

Thanks,
Prasad
_______________________________________________
fcoe-devel mailing list
[email protected]
http://lists.open-fcoe.org/mailman/listinfo/fcoe-devel

Reply via email to