RE: SCSI URL scheme [WAS: Re: iSCSI: 2.2.6. Naming & mapping]

[julian_satran@il.ibm.com]

Bob,

The trouble we had with numbers is that they tie you up to a specific
addressing scheme
and not having any hierarch they don't scale well. In a contained
environment that is fine
but beyond it (and we are there) a flat numbering scheme is bad.

I did not want to enter this territory that early but we are forced to
choose between
IPV4 and IPV6 - and we choose not to choose. Its that simple!

Julo

Robert Snively <rsnively@Brocade.COM> on 28/09/2000 20:04:55

Please respond to Robert Snively <rsnively@Brocade.COM>

To:   Julian Satran/Haifa/IBM@IBMIL, Douglas Otis <dotis@sanlight.net>
cc:   Daniel Smith/Almaden/IBM@IBMUS, ips@ece.cmu.edu
Subject:  RE: SCSI URL scheme [WAS: Re: iSCSI: 2.2.6. Naming & mapping]




Julo,

I am sorry, but I disagree with you both practically and philosophically.
Textual names are practically useless as identifiers, as witness the
challenge of finding "Bob Smith" in an American phonebook.  Every
time you walk into a store where your name is kept in a data base,
they ask for your phone number, not your name.  Your passport,
credit card, driver's license, and all other useful identifiers
use a registered number set that allows for no (legal) duplication.

SCSI has taken the same approach with LUN identification, recommending
one of several registered numbers as the sole and immutable designation
of the logical unit.

Fibre Channel and Ethernet have taken the same approach at the
Port/Node WWN and MAC address level.

Use of a URL is a virtualization of the underlying name structure
and should not be implemented except as a temporary convenience for the
highest level client who is constrained to a very small search
subset by his login authorizations.  The virtualization is always
at risk of security breaches.

In the storage environment, the underlying structure is of vital
interest.  Any given data is present on specified physical and
logical units and consistency of the data is maintained on the
basis of an intermediate view of the lower level addressing structure
which must also have the proper uniqueness guarantees.

Address it with a URL on the browser if you wish, but understand that
you are talking to a data block on a single storage unit attached to
a network server having a specified IP address mapped to a
registered MAC address.

Bob

>
>  There seem to be practical and philosophical issues that you
>  consistently
>  ignore.
>  Textual names are more stable than underlying network
>  addresses and thus
>  better
>  for applications (I assume that your card starts with your
>  name not phone
>  number and
>  that the former has changed less than the later).
>  I the iSCSI realm - and that is the only one under scrutiny
>  - a pertinent
>  observation was
>  made by David Black a century ago during our (too) long
>  naming discussions
>  - that a scheme that will allow the third party commands to
>  make the name
>  to address conversion
>  at use time is better than one that requires this mapping to
>  be made ahead
>  of time.
>  Unfortunately that is not yet possible with SCSI but we are
>  informed that
>  it soon be.
>  And for authentication the textual address in login is yet
>  another mean to
>  prevent
>  errors and/or unauthorized access if the target chooses to use it.