Re: Older StorageWorks Parts Not Available
From: David J Dachtera (djesys.nospam_at_comcast.net)
Date: 01/01/05
- Next message: David J Dachtera: "Re: Migration from SCSI to EMC"
- Previous message: Phillip Helbig---remove CLOTHES to reply: "(DIS)MOUNT/POLICY=MINICOPY in mixed VAX-ALPHA cluster"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Sat, 01 Jan 2005 13:54:45 -0600
"Main, Kerry" wrote:
>
> > -----Original Message-----
> > From: David J Dachtera [mailto:djesys.nospam@comcast.net]
> > Sent: December 30, 2004 6:24 PM
> > To: Info-VAX@Mvb.Saic.Com
> > Subject: Re: Older StorageWorks Parts Not Available
> >
>
> [snip]
>
> > Centers which are within the same metropolitan area have no business
> > using anything other than HP field service to deliver parts. For
> > example: (break out your maps) Delivery from Chicago to Oak Park (less
> > than 8 miles from the lakefront to Oak Park's western-most border),
> > easily walkable within a day, drivable within two hours even in the
> > worst weather conditions) still takes five days via UPS standard
> > (surface) service.
> >
>
> Are you saying that it took 5 days to receive a part that was in the
> same city??? I have a hard time believing that.
In keeping with my newsgroup policy of "prove me wrong", try it for
yourself. Ship ANYthing from your office to your home via UPS standard
surface delivery (pka "brown label") - a box of staples, and old, broken
stapler, a roll of tape, whatever - and see how long it takes.
...take pictures - all sides - of the package leaving your facility and
when it arrives at its destination, then compare.
In our case, the hardware ticket was opened on a Monday. The FSE brought
a part on site that day (DOA), parts were (re-)ordered on a Tuesday, and
arrived the following Saturday evening. In the meantime, the FSE brought
two more DOAs at 36-hour (approx.) intervals, for a total of five DOAs.
> By the way - most Field Service companies these value their technicians
> time much more than a delivery van guy i.e. it is better to have the
> tech work on problem, talk to Customer, whatever.. As opposed to waste
> tech time doing low level task like driving across town, picking up
> part, finding parking, signing back in again etc etc. Heck, even Digital
> was doing this back when I was in CS Support.
Really? For as long as I can remember, the FSEs have been siging out
parts before leaving for the customer site and bringing the parts with
them, if the FRU was known at that time, and once on site it *NEVER*
took more than 18 hours to get a part delivered (except for a "recent"
event with a GS160 part replacement which took over 26 hours - FSE was
on-site the entire time), less for life-critical systems (as are
typically found in heatlthcare).
> > > As to the DOA's, that obviously needs to be looked into to determine
> > > what happened. Typically what gets looked into are such things as :
> > > - Was it bad from the center?
> >
> > Good bet.
> >
> > > - Was it packed
> >
> > Yes.
> > > and handled correctly?
> >
> > Cannot be gauranteed. UPS is infamous for "package volleyball"
> > tournaments within their truck depots.
> >
> > > - Was the installation of the part done by a qualified
> > Field Engineer?
> >
> > Not a factor.
> >
> > > (need to consider things like static issues, quiescent
> > buses properly
> > > before replacement, proper commands to be executed from
> > disk controllers
> > > like HSx etc. )
> >
> > Must be done during uptime. "Quiescing", etc. is not possible.
>
> Who did the parts replacement - CS or the Customer?
>
> If the disks were replaced by someone who did not follow the recommended
> replacement procedures - which on some controllers do require bus
> quiescing (can be done on line at less busy time of day as I recall). On
> HSx controllers, you do also not simply swap bad drives with good drives
> - there are specific HSx processes to move drives in and out of failed
> sets etc.
Here again, and as noted earlier, all documented procedures were
followed by qualified personnel. The parts were DOA. Period, end of
statement.
> >
> > > - Was the revision of the part an issue? Some new
> > equivalent parts might
> > > have HW/SW compatibility issues with older systems that need to be
> > > addressed before they can be installed properly.
> >
> > Should have been addressed before the part was shipped.
>
> So, tell me how the logistics person is supposed to know what firmware
> you are running or what OS version / patch level you have in place.
Thank you for underscoring my point so well. A "logistics person" has no
business being involved at that level (or any other, really, in this
scenario).
> Some
> new parts have incompatibility issues with older OS versions, firmware
> etc. CS folks know these things and that is one of the reasons why for
> mission critical environments, one should have qualified CS folks doing
> the replacement.
...and selection of parts to be shipped, not a trained UPS ape.
> Now, perhaps CS did the replacements in your case, but I have seen more
> than a few cases where Customers think they can simply swap out the
> brick drives without understanding that there is more to it.
See the above.
> >
> > > - Has this part a history of problems or was it only at
> > this one site?
> >
> > Multiple targets - question irrelevant.
> >
> > > - What failure symptoms were seen i.e. were they the same
> > for each DOA
> > > or were they different?
> >
> > All mostly different. Some never spin up, others spin up then
> > die, some
> > start to work then fail, ...
> >
> > > - other issues looked at as well ...
> >
> > ...and all covered in great detail. The parts were DOA. Period.
> >
>
> I am not saying they weren't, but getting that many bad DOA's of the
> same part seems awful strange to me.
To me, as well. Suggests a new height in the field of technical
incompetence.
> As a former CS support resource, while one can obviousl;y not rule out
> some bad parts until you investigate it further, I would initially be
> looking for some common threads besides just immediately blaming the
> Logistics center or those delivery guys who "are known for having volley
> ball tournaments with packages".
Care to elaborate on that? Disk drives are designed to survive certain
G-force (impact) levels. What part of "package volleyball" suggests
proper handling of fragile components?
(By the way, "package volleyball" within UPS terminals has been
documented by the media, and has even been admitted by UPS itself. So, I
will discount any claims of "anecdotal" on that score.)
> Improper eco levels (disk or controller), improper replacement
> procedures, not allowing disks to thermally stabilize after being
> outside in truck for awhile, not being aware of latest known issues with
> that particular disk and/or controller etc would all be areas to
> investigate.
See the above.
-- David J Dachtera dba DJE Systems http://www.djesys.com/ Unofficial OpenVMS Hobbyist Support Page: http://www.djesys.com/vms/support/ Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/ Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/
- Next message: David J Dachtera: "Re: Migration from SCSI to EMC"
- Previous message: Phillip Helbig---remove CLOTHES to reply: "(DIS)MOUNT/POLICY=MINICOPY in mixed VAX-ALPHA cluster"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|
