> I made the assumption, a false one from looking at the backtrace,  
> that I/O was occurring to the device. When thick_provo_start() is  
> running the device is marked as offline so that initiator can't do  
> anything other than send a REPORT_LUN command to the
> device.

@rutger (Rutger Bevaart):  how man iscsi targets were defined and
in active use when iscsitgtd was crashing?


Maybe at least one iscsi target was in use, while a new one was
created?  iscsitgtd crashed while creating a new target, and this
might result in the iscsi initiator timeouts for the other (active) iscsi
targets?


> I'll have to look at that offset and find out what's happening. That  
> should be easy to trace.

The crash seems to happen at line 1392 (or 1394?) in
usr/src/cmd/iscsi/iscsitgtd/util.c

  1392  free(tp->targ_name);
  1393  queue_free(tp->q, NULL);
  1394  free(tp);
 
 
This message posted from opensolaris.org
_______________________________________________
storage-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/storage-discuss

Reply via email to