Conference 7.286::digital

    
    The first thing you have to do is make sure everyone is talking about
    the same "problem". .0 doesn't seem to have a clear reference to the
    problem -- is it that there is a bug encountered when a certain drive
    is mated with a certain controller, or is it that this combination has
    just not been qualified? (And how does one define the term "qualified"?)
    
    Experience tells me that unless you have a clear, concise,
    reproducible, *WRITTEN* and universally referenced (eg, via IPMT case #
    or equivalent) description of the "problem" you need solved, you're just
    mixing mortar for the Tower of Babel.
    

The problems, as I see them, are manifold:

	there is a bug or incompatibility between the RZ28M firmware and
	the KZxSC controllers.  I suspect somebody's working around the
	clock to fix this.
	
	we allowed ourselves to be a position where we don't have a 2GB 
	drive-based, PCI/EISA controller-based RAID solution.  Didn't we know,
	back in the spring or early summer when the RZ28 drives started
	running out, that we'd need a replacement?

	if we're shipping the SD001-CA and SD001-DA, which consist of a one-
	or three-channel KZPSC and 2GB drives, is it with with RZ28Ms?

	if "conditions beyond our control" caused us to run out of the RZ28
	drives before we had time to qual another drive, why didn't have a 
	plan to do *something* about it?

	if there is a plan, why don't we know about it?

	there are many situations where our customers are trying to use RZ28M
	drives on the these two controllers, because we sold it that way

	evidently, the RZ28D has been tested or qualified with the controllers
	but has not been qualified in any systems


Bruce

    I think these discussions should be held in the
    MSGAXP::STORAGE_PRESENTERS conference.  Moderated by SHAKEY::SENGUPTA.
    
    This is the Storage PID Presenters conference, a restricted conference,
    so you need to find out the requirements for entry.

RE: .4 Why in heavens name would you suggest that the discussion of a potentially critical issue 
such as this one be moved to a restricted conference. The base noter, while admittedly also 
venting a bit was, IMHO, trying to alert other unsuspecting Digits about this problem. What 
possible good would it do, other than to bring it again to the attention of people who undoubtedly 
know about this problem, to move it to a restricted Storageworks presenters conference. That would 
certainly defeat one of the basic purposes of these conferences. Hopefully, public airing of our 
own internal stupidity contributes to the process of identifying and weeding out the folks who 
make some of these stupid decisions. 

-OFWAMI-

   <<< Note 4241.5 by NQOS01::nqsrv326.nqo.dec.com::Werner "NORMAN WERNER" >>>
                      -< Don't hide the smoking guns... >-

  RE: .4 Why in heavens name would you suggest that the discussion of a
  potentially critical issue such as this one be moved to a restricted
  conference. The base noter, while admittedly also venting a bit was,
  IMHO, trying to alert other unsuspecting Digits about this problem.
  What possible good would it do, other than to bring it again to the
  attention of people who undoubtedly know about this problem, to move
  it to a restricted Storageworks presenters conference. That would
  certainly defeat one of the basic purposes of these conferences.
  Hopefully, public airing of our own internal stupidity contributes
  to the process of identifying and weeding out the folks who make
  some of these stupid decisions. 

  -OFWAMI- 

I have more to do during the day than monitor this soapbox, BUT someone sent me
this note thread VIA mail and I had to put in my .02 worth.


>    I understand the required firmware fix for the RZ28M is one to two
>    months away.  Don't know when testing and support will be in place.....

Gee could have fooled me. Revision 568 was cut into manufacturing the end of
August. 616 is still in qual but has no fixes that preclude it's use. Only 1 fix
in that code that I am aware of and it's a minor one.

>    Only place I've actually seen info on this drive problem is in notes.
>    .................

You and me both and ALL RZ28M IPMT cases come through ME!

I'm sure product managers will be adding more responses to this as this is not
correct information being passed around.

Roger.

    Werner describes my motivations well.  I was venting, but mostly
    because I *know* that there have to be cases out there where customers
    have RZ28M drives and KZxSC controllers.  And friends, the combination
    doesn't work.
    
    Look in the current SOC.  In the AlphaServer 1000, 2000, 2100 sections
    we say, in adjacent configuration steps, to use the KZPSC or KZESC and
    the RZ28M as the 2GB drive of choice.  
    
    In the Sales Update article announcing the RZ28M and the RZ26N, right
    at the top of the article, we say the RZ28M replaces the RZ28.  Only
    when one looks almost at the bottom of the article does one find the
    indirect statement of non support "Support for the KZESC and KXPSC [sic] 
    RAID controllers will be included in futre [sic] high performance 
    StorageWorks disk drives."
    
    Who reads Sales Updates, anyway?
    
    The point is that it *should* work and it doesn't.  Okay?
    
    Now, it's being worked, I hope.
    
    But I'll ask the question I asked in the base note.  Does anyone, who
    can do anything, understand this problem from (a) the customers' 
    pespective and (b) the systems' perspective?
    
    Re: the RZ28D...
    It's not on the latest (September) Supported options list,
    http://www.service.digital.com/alpha/server/as2100/docs/
                              supported_options_abs.html
    
    Bruce
                              

    
    If the Customer configuration is not working the Escalate it to
    engineering. Don't hope it is being worked. Drives are being released
    so fast that you cannot always trust the docs. The RZ28M is supported.
    
    Roger Patenaude
    Storage External Products
    Continuation Engineering
    

    I've been forwarded the base note and some of the replies of this note;
    due to a week at COMDEX this is my first opportunity to reply.
    
    The configuration of RZ28M drives and the KZESC or KZPSC is supported
    on the AlphaServer 1000 and 400 systems.  I have had NO problem reports
    forwarded to me and we've had no IPMT cases come in through the
    customer support side.
    
    It seems to me that the only problem that's actually been reported here
    to date, is that the DOCUMENTATION hasn't kept up with reality...
    
    If anyone has experienced a functional problem with these combinations,
    please let us know.
    
    

RE -1

>>  It seems to me that the only problem that's actually been reported here
>>  to date, is that the DOCUMENTATION hasn't kept up with reality...

Probably because the tech writer responsible for the DOCUMENTATION got 
outsourced!