Conference ssdevo::hsz40_product

757.0. "battery problems" by TOPTEN::KELLER () Mon Feb 03 1997 12:57

			(Posted in HSJ/HSZ notes file.)

Could someone shed some light on this, please?

  a. Is there any relief in sight for the battery problems that we have with
  HSJ40's and HSZ40's?  Currently I have about a 40% failure rate.

  b.Document AV-QPDXA-TE (Array Controller Cache Battery Replacement Letter)
  states that if you have redundant controllers with Raid or Mirror sets and
  need to replace the battery in one of the controllers, you need to shutdown
  both controllers.  Isn't this contrary to one of the reasons as to why there
  are "redundant" controllers?  Almost a year ago, when the Rev B01 battery to
  end all problems came out, I replaced 12 or so online with no problems by 
  running the current version of C_SWAP (HSJ40's running V2.7 I think).  At 
  the time we had several Raid sets on those controllers.  Since then we have
  gotten away from Controller based raid and we are running Host based raid
  so it is not an issue.  However I now have several HSZ40's with Raid and 
  some mirrowing which according to the procedure, I have to bring down the
  controllers/system.  Is this correct, am I reading this wrong, is this a 
  temperary problem waiting for a solution or what?  I am currently running
  V2.7-1 on the J's and 3.0-2 on the Z's.

This could potentially impact a lot of systems/databases that are required to
be up 24 x 7.


Regards,

Larry

    I think the reason for shutting down both controllers was to replace
    both sets of batteries, If you are sure that the batteries are good
    on one of the controllers and you are replacing only one set, then
    the C_swap could be used.  However be positive that the other
    controller's batteries are good.