Re: iSCSI: Concerns raised about ACA and denial of service conditions

["Julian Satran" <Julian_Satran@il.ibm.com>]

Ralph,

I need some more clarifications.
If ordered tasks between initiators have to be coordinated then what you
view as
a DOS attack is perhaps desirable behavior (everything goes down the drain
because
we want coordination); the only open issue is why can't any initiator - in
this case - clear the ACA?

If they don't have to be coordinated then we don't have a problem - right?

Can I call you somewhere?  Today after 12 AM CDT or tomorrow after 10 AM
CDT ?

Thanks,
Julo



Ralph Weber <ralphoweber@compuserve.com> on 24-07-2001 16:46:27

Please respond to ENDL_TX@computer.org

To:   ips@ece.cmu.edu
cc:
Subject:  Re: iSCSI: Concerns raised about ACA and denial of service
      conditions




Julian,

> Isn't the behavior you describe being controlled by the
> TST (and QErr) field from the Control Mode Page (SPC2)
> and the value of 001 insures that an ACA condition caused
> by one initiator does not cause tasks from another
> initiator be rejected?

The behavior I described assumes TST=000, which I consider
to be the default case.  QErr affects whether pending tasks
are terminated when an error occurs, so I do not think QErr
applies here.

Note that TST=001 affects more than just the error behavior.
If TST=001, then ORDERED tasks from one initiator are not
coordinated with ORDERED tasks from other initiators.  ABORT
TASK SET only the tasks from one initiator making it the
same as CLEAR TASK SET.  These behaviors could be viewed as
undesirable.

Thanks.

Ralph...