Re: iSCSI: synch and steering comments

[julian_satran@il.ibm.com]

Answers in text. Thanks, Julo

"Mallikarjun C." <cbm@rose.hp.com> on 27/03/2001 19:44:43

Please respond to cbm@rose.hp.com

To:   ips@ece.cmu.edu
cc:
Subject:  iSCSI: synch and steering comments




Julian,

Following are the comments I have on Synch and Steering in rev05.

-Suggest replacing "it" in the following sentence in section 1.2.8.2, page
23
 (first sentence in a new para):

 "According to our model of layering, iSCSI considers the information
    it delivers (headers and payloads) as a contiguous stream of bytes
    mapped to the positive integers from 0 to infinity."

 with "that a Synch and Steering layer" to make it clear.  The sentence
 is ambiguous about the direction of delivery with the way it is.

+++ I adjusted to clarfy direction ++

-Suggest adding the following statement to section 1.2.8.2.

 All conventional, in-order data arrival notifications generated by TCP
 are passed through to iSCSI by the Synch and Steering layer after
 appropriate data placements while none of the out-of-order data placements
 that it performs are communicated to upper layers.

+++ I have added the following to 1.2.8.2

   On the incoming path the Synch and Steering layer does not change the
   way TCP notifies iSCSI about in-order data arrival.  All out-of-order
   data placements
   performed by the Synch and Steering layer are hidden from iSCSI.

   I have aloso changed a bit the figure to convey better the fact that TCP
   and Synch&Steering are related (not strictly layered +++

   ++++

-Section 1.2.8.2 states that a Synch and Steering layer is optional.
 It has to be qualifed that it is optional only for those iSCSI devices
 which perform connection recovery on header digest errors, since that's
 how they cope with loss of framing. (I guess this may change in next rev?)

+++ with the new format I think that we have:

- one more chance if we go for format 1 or
- drop the connection on header error

In both cases we can leave synch and steering optional

+++

-It appears to me that at least one Synch and Steering layer must be
 defined/referred to as the minimal implementation in the main draft to
 enable interoperability, when implementations do implement Synch and
Steering.

+++ why ? +++

-I am somewhat confused about the following statement in section 1.2.8.2:
    "The Synch and Steering Layer is required to add to every sent data
    item (IP packet, TCP packet or some other superstructure) enough
    information to enable the receiver to steer it to a memory location
    independent of any other piece. "

 Clearly from the way I understood the markers in Appendix.C, it doesn't
 comply with this requirement.  A more generic statement would be:
    "The Synch and Steering Layer is required to add adequate
    information to the data stream to enable the receiver to quickly
    steer the stream to its final memory location, even in the face of
    discontiguities in the stream. "

+++ Markers are but one example and have only the Synch component. The
statement refers to a full steering (RDMA) scheme +++


--
Mallikarjun


Mallikarjun Chadalapaka
Networked Storage Architecture
Network Storage Solutions Organization
MS 5668   Hewlett-Packard, Roseville.
cbm@rose.hp.com