Hang problem on write in ST_TTS4:Receive_Data_Out state

George Penokie gop at us.ibm.com
Fri Apr 4 09:39:33 PST 2003

* From the T10 Reflector (t10 at t10.org), posted by:
* George Penokie <gop at us.ibm.com>

There is a condition that will cause a target to hang in the
ST_TTS4:Receive_Data_Out state. This occurs if a XFER_RDY frame is
successfully sent and there is a legal connection close but the Initiator
never reestablishes a connection for that I_T_L_Q nexus.

I do not believe we should leave the recovery of this up to the imagination
of the implementers when we have a mechanism defined that fits this
condition. What the ST_TTS4 state should have added to it is an I_T nexus
lost timer. It would be initialized to the value in the I_T nexus loss time
field in the Protocol-Specific Port Control mode page and started on entry
|from the ST_TTS2:Send_Data state or the ST_TTS6:TTS6:Process_Data_Out state
and reset on receipt of a Data-Out Arrived message. If the timer times out
before all the data indicted in the XFER_RDY is received than a this state
would send a Nexus Lost confirmation to the SCSI application layer. This
state machine would then terminate after sending the confirmation.


Bye for now,
George Penokie

Dept 2C6  114-2 N212
E-Mail:    gop at us.ibm.com
Internal:  553-5208
External: 507-253-5208   FAX: 507-253-2880

* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org

More information about the T10 mailing list