Importance: Normal
MIME-Version: 1.0
Sensitivity: 
To: Caitlin Bestler <cait@asomi.com>
Cc: wrstuden@wasabisystems.com, ips@ece.cmu.edu, rddp@ietf.org
Subject: Re: iSCSI/iWARP drafts and flow control
Message-ID: <OFA45B4A51.40A3A6A4-ON87256D83.0083CD9E@us.ibm.com>
From: Mike Ko <mako@almaden.ibm.com>
Date: Fri, 15 Aug 2003 18:04:08 -0700
Content-Type: text/plain; charset="us-ascii"
Sender: owner-ips@ece.cmu.edu
Precedence: bulk

As I indicated to Caitlin in an earlier e-mail, since each iSCSI 
control-type PDU from the target to the initiator is associated with a 
StatSN, the ExpStatSN from the initiator can be used to advance the 
"credit" at the target.

The question is for iSCSI control-type PDUs from the initiator to the 
target, what is the mechanism for the target to advance the "credit" at 
the initiator?  For PDUs where the Immediate flag is not set, the CmdSN 
and MaxCmdSN mechanism in iSCSI are already available and the new "credit" 
scheme is not necessary.  The unsolicited SCSI Data-out PDUs are covered 
by FirstBurstLength and again the new "credit" scheme is not needed.  The 
"credit" scheme is needed for PDUs where the Immediate flag is set or the 
SNACK Request (note that SNACKs are not needed in iSER).  For these PDUs, 
the target may not be able to advance the "credit" at the initiator 
individually.  One possibility is to wait for the initiator to issue an 
iSCSI control-type PDU with a new CmdSN but without the Immediate flag, 
and if the target responds with an iSCSI control-type PDU with ExpCmdSN 
set to CmdSN + 1, then "credit" at the initiator is restored to its full 
negotiated (or by definition) value. 

As Caitlin pointed out, since no one has yet identified a specific need 
for per-session negotiation of these credits, a simple rule enumerated in 
the draft may be sufficient. 

Mike Ko
IBM Almaden Research
mako@almaden.ibm.com

To:     wrstuden@wasabisystems.com
cc:     Mike Ko <mako@almaden.ibm.com>, ips@ece.cmu.edu, rddp@ietf.org 
Subject:        Re: iSCSI/iWARP drafts and flow control




On Friday, August 15, 2003, at 04:12 PM, wrstuden@wasabisystems.com
wrote:

> Sorry for my delay in responding.
>
> One of the things I've been wondering through this whole thread is
> can't
> we expand iSCSI negotiation to handle these parameters (like how many
> async messages are ok, how many immediate commands, etc.)? That way, to
> the extent that iSCSI/iWARP needs new parameters, we can decide on them
> out easily.
>
> Maybe it's been in everyone's thoughts, but the thread had been talking
> about needing new limits and talking like it's hard to do. Can't we
> just
> negotiate them?
>
> Take care,
>
> Bill

Bill,

There are two separate issues.

The first is establishing what the credit limits are. That could
be negotiated on a per session basis, or it could be established
by the draft as a simple rule. You are correct that negotiating
is not terribly complex. Therefore, if there is *any* benefit to
varying those limits on a per-session basis, then negotiations
should be allowed.

So far, however, I do not believe that anyone has identified
a specific need for per-session negotiation of these credits.

The second issue is the mechanism by which credits are replenished.
The iSER draft authors were opposed to creation of additional
messages to exchange credits. My contention is that under virtually
all circumstances, credit replenishment can be inferred from
existing message exchanges. Under worst case circumstances a
"NOP" might have to be sent that would not have otherwise
been required. No additional message types are required.

I believe the last few prior exchanges may have established
that tracking credits is not as onerous as the draft authors
had originally thought. But it is on that issue, not the
overhead of negotiation during setup, that was their concern.





Caitlin Bestler - cait@asomi.com - http://asomi.com/