I respect your opinion, Bill, but I feel you don’t have all the facts:
You wrote it is not “officially” supported by HP. Duh!! How many third-party
disk subsystems does HP support? It is political. Here we have a disk
subsystem that can potentially double or triple the workload, and prevent the
customer from doing an unnecessary CPU upgrade. It also provides other
advantages over JBOD and EMC.
>At the very least, David's company should get HP to say SSA does no harm
>and is at least as robust as JBOD (Just a Bunch Of Disk).
I want that too. HP doesn’t return our calls. For those of us that went to
HPWorld97 open/executive forum, they said “they would look into it”. They
haven’t! In defense of HP, I have exchanged several emails and telephone
conversations with Scott McCullen, out of the CSY lab. He gave me another name
last night to contact. Perhaps she will be responsive. I’ll still give HP the
benefit of the doubt, and offer that we haven’t been talking to the right
people <sigh>
We have offered to supply everything they need to run their tests, and even pay
HP for their time in voicemail and email … they never responded.
HOW WE HAVE TESTED SSA TO DATE:
* I hooked up HP 3000s with both the HSC and HP-PB bus to a SCSI analyzer,
With both HP JBOD and our SSA drives on the back end. The ONLY difference
Is in the timing, where we completed the I/O operations much quicker. Some
of
the operations actually returned in nanoseconds, rather than milliseconds.
* HP engineers have done the complete series of diagnostic programs at both our
Site and end-user sites, without problems. Believe me, they were looking for
a way
To create a data loss, or problem, as these were trial/demo situations. No
luck on
HP’s part --- no problems found. {No, actually we did find one with
hot-swapping,
and did a firmware update around December}
We have also actually have drives become sick or outright fail. HP’s
predictive saw
the problem, and the customer got notified in the same way that they would if
it was
an HP disk drive. In fact, the HP salesrep - who will remain anonymous -
commented that it wasn't fair that predictive should even work with IBM disk
drives. The customer then promptly reminded him about antitrust laws.
We also have our own diagnostic program which can show additional information
that
HP can’t address with their diagonstics. We can do things such as:
* Turn write cache on/off at the drive level
* Detect a failure in a cable or power supply
* Make drives invisible to a specific adapter
So, if there is a potential for problems, I submit that the customer would have
the
rame risk with HP drives, as they behave “exactly” the same. (Only we have
more
features, and more redundancy). There are also some additional enhancements
coming soon which will knock your socks off … such as getting beyond the
255 disk drive limit!
Bill, how do I get in contact with this executive forum on high performance???
David A. Lethe
[log in to unmask]
(972) 208-3660
(remove the XYZ to send email to me)
Bill Lancaster wrote:
> Rick wrote:
> At 04:06 AM 3/18/98 GMT, David A. Lethe wrote:
> >On Tue, 17 Mar 1998 10:41:51 -0500, Rick Tomlinson <[log in to unmask]>
> >wrote:
> >
> >>Has anyone ever used or heard of anyone using mirrored EMC Symmetrix
> >>drives for the system volume set? What other alternatives are there for
> >>adding this type of redundancy?
> >>
> >>
> >>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> >>Rick Tomlinson [log in to unmask] (336) 758-5811
> >>Information Systems fax (336) 758-7127
> >>Wake Forest University www.wfu.edu/~rick
> >>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
> And then David wrote:
>
> >Look at IBM's SSA for the HP3000. It is typically several HUNDRED
> >PERCENT faster, at about half the price. Hardware mirroring is in
> >beta now. All the usual features you would expect, including:
> > - hotswap
> > - multiple initator support (with limitations)
> > - RAID capability
> > - Works with HP's predictive & other diagnostics
> > - priced around 50 cents/MB.
> > - very high density .. almost a terabyte in a HP 19" rack
> > - no single point of failure (redundant paths, cables, power)
> > - add, delete, reconfigure, mount them to different systems without a
> >reboot!!! (requries 5.5)
> > yada yada yada
> >
> >I know, I was the project leader on the team that ported it to the
> >HP3K I contracted Michael Hensely with Allegro to port the diagnostic
> >routine to MPE/iX ... So you know it was done right!
> >
>
> Just a couple of comments. First, in direct answer to Rick's question, you
> can use EMC Symmetrix drives for the system volume set but only in a RAID 1
> configuration. Works great.
>
> Second, a little caution is in order before making the jump to SSA for
> production. I'm sure that David's company did a great job in bringing the
> technology to the 3000. My concern is that it is not "officially"
> supported by HP. Until HP gets on the bandwagon of SSA on the 3000, I will
> find it very difficult to recommend it in the 3000 environment. At the
> very least, David's company should get HP to say SSA does no harm and is at
> least as robust as JBOD (Just a Bunch Of Disk).
>
> Lastly, just because Allegro ported the diagnostic routine doesn't
> necessarily mean that 1) SSA was "done right" on MPE/iX and 2) that Allegro
> endorses the technology for the 3000.
>
> There are other options for this type of redundancy:
>
> 1. Disk arrays for the system volume set
> 2. Mirrored Disk/iX for all production volume sets
> 3. Quest Netbase/SharePlex for full system redundancy
>
> (pick any or all of the above)
>
> BTW, Interex have recently formed a new High Availability Forum which will
> address high availability issues across all HP product lines. Isaac Blake
> and myself are co-chairing the Executive Committee for this Forum. We
> invite you all to give us some feedback on what high availability issues
> are important to you and look forward to representing you to HP. We will
> be having several meetings this year to begin the process of advocacy in
> the high availability arena so look upon this time as ripe opportunity to
> get your requests known to HP in a fully supported way.
>
> Bill Lancaster
> [log in to unmask]
|
