[comp.periphs.scsi] SCSI monitors

km@mathcs.emory.edu (Ken Mandelberg) (07/02/90)

I saw a scsi monitor advertised for about $850. This is a box that
plugs into the scsi bus and displays some info on a built in lcd
screen. At very least it shows the last command, I'm not sure what else
from the ad. They do show some pictures of the box monitoring an
exabyte, where the lcd shows transfer rate, error rate, and space
remaining.  In fact the display looks a lot like the built in lcd that
some exabyte resellers include in the box.

Can anyone comment on how useful this sort of scsi monitor is? It
strikes me that it would help diagnose some gross problems, but would
be pretty useless with problems that occur in the middle of a lot of
traffic.

At half the price, to me it would be a clear buy, even if I just used
it to monitor an exabyte most of the time.  At $850 it takes a bit more
thought, which is why I'm asking for feedback. In fact if anyone knows
of a lower cost unit, I would like to hear about it.

If your wondering why I haven't mentioned the company name, its nothing
subtle. The ad is not where I can get to it at the moment.
-- 
Ken Mandelberg      | km@mathcs.emory.edu          PREFERRED
Emory University    | {rutgers,gatech}!emory!km    UUCP 
Dept of Math and CS | km@emory.bitnet              NON-DOMAIN BITNET  
Atlanta, GA 30322   | Phone: (404) 727-7963

andrew@alice.UUCP (Andrew Hume) (07/10/90)

this group may have covered this before but can anyone point me towards
a sophisticated SCSI monitor? i had in mind something i can plug into
a scsi bus and display for me the last many (10-100s) commands and responses.
price, for now, is no object.

andrew@research.att.com

garyb@SSD.CSD.HARRIS.COM (Gary Barton) (07/11/90)

In article <11038@alice.UUCP> andrew@alice.UUCP (Andrew Hume) writes:
>
>
>this group may have covered this before but can anyone point me towards
>a sophisticated SCSI monitor? i had in mind something i can plug into
>a scsi bus and display for me the last many (10-100s) commands and responses.
>price, for now, is no object.

We have a couple of Pacific Electro Data PED4000 series of PC
expansion cards and software.  I sometimes wonder how we got by before
we had these things.  If all you want is a passive analyzer, ask about
the PED 4003 state and phase analysis package.  If you require target
emulation as well, then ask about the 4033 model.  To date, the only
problem we have had with these boards is that for high BW transdfers,
the data acquistion board is a little too slow.  This results in
garbled data showing up for faster peripherals (eg. sync xfers).  This
hasn't been a problem since all other bus phases are handled
correctly, and you usually don't care to see all the data anyway.
This board requires one full length 8 bit slot of an IBM PC compatible.

PED has new product they are developing, the PED 4500 series.  Because
of the data BW problem, I really wanted the new faster product, but
the target emulation package was not yet ready.  So they allowed me to
buy the 4033, and upgrade to the 4533 when it's ready, with a 100%
trade in credit.  This should be out in the near term (last schedule I
heard was sometime in July).

Anyway, here's their address and phone for those interested:

	Pacific Electro Data, Inc.
	14 Hughes, Suite B205
	Irvine, CA  92718
	(714) 770-3244

Here are some of the prices I was quoted back in January.

	PED4003	 ~$2000  Old product, analyzer only
	PED4033	  $3495  Old product, analyzer and target emulation
	PED4503	 ~$5000  New product, analyzer only
	PED4533	  $6995  New product, analyzer and target emulation

cgn@leo.UUCP (Chris Nieves) (07/12/90)

In article <11038@alice.UUCP>, andrew@alice.UUCP (Andrew Hume) writes:
> 
> 
> a sophisticated SCSI monitor? i had in mind something i can plug into
> a scsi bus and display for me the last many (10-100s) commands and responses.
> price, for now, is no object.
> 
> andrew@research.att.com

We've been using the Ancot analyzer for a year or so.  If price is really
not a concern you should give them a call:
            Ancot
            1755 E. Bayshore Rd
            Suite 18A
            Redwood City, CA 94063
            (415) 363-0667

Their analyzer can monitor the bus and capture upto 32K worth of trace data
at 50ns resolution with time stamping.  They also can be either an initiator
or target on the SCSI bus.  We've found it really handy in debugging our
host adapters and getting performance numbers on disk drives.


-------------------------------------------------------------------------
Chris Nieves
UUCP : ccicpg!leo!cgn  or cgn@leo.ccicpg
USPS : ICL North America, 9801 Muirlands Blvd., Irvine, CA 92718-2521
PHONE: (714) 458-7282
-------------------------------------------------------------------------

PS-I'm not associated with them at all...just like their product.

paul@ppgbms (Paul Evan Matz) (07/13/90)

In article <338@travis.csd.harris.com> garyb@SSD.CSD.HARRIS.COM (Gary Barton) writes:
>In article <11038@alice.UUCP> andrew@alice.UUCP (Andrew Hume) writes:
>>
>>this group may have covered this before but can anyone point me towards
>>a sophisticated SCSI monitor? ...
>
>We have a couple of Pacific Electro Data PED4000 series of PC
>expansion cards and software.  I sometimes wonder how we got by before
>we had these things....
>
>...Anyway, here's their address and phone for those interested:
>
>	Pacific Electro Data, Inc.
>	14 Hughes, Suite B205
>	Irvine, CA  92718
>	(714) 770-3244
>
I just wanted to second this recommendation and add a little.  We've been
using their 1st generation hardware and software for almost 2 years without
any reliability problems.  Initially, when we uncovered a few problems with
their emulator software, their tech support was instantaneous;  All information
exchange was done via fax, and software fixes were sent overnight express
(not to imply that you'll need anything fixed at this point).

They have both state and phase analysis modes, as well as both target and
initiator emulation.  Although we've never done it, target/initiator emulation
and one analysis mode may run simultaneously.

The emulation is done by writing in a C like programming language that is run
by an interpreter;  The interpreter is the guy who controls the SCSI hardware
during emulation.  The analysis hardware is also programable, although the
language is quite cryptic, due to the fact that it is run by a little bit-slice
processor who executes one instruction every 100 nanoseconds (PED-4001 hardware).
You get 32 words worth of program, which turns out to be enough for most
applications.

The emulation on a PC/XT class machine is somewhat slow.  I don't know what
kind of inprovements can be attained by going to a faster PC, or how much
faster their next generation of hardware is.

The documentation is a bit obtuse;  You can tell it was written by the guy
who designed the system.  Once one wades thru it, though, the result is
being able to simulate just about anything you want, and being able to
look at it in detail.  The capture buffer on the 4001 is 2K words;  If
you set up your analysis to ignore most of the data phase, you can grab
20-30 full transactions which are all timestamped relative to each other.
Triggering the capture is configurable and can be based on things like
which intiator is talking to which target, what the command being sent
is, etc..  (This is when running their state analysis mode of caputure).
I've never run their phase analysis, so I can't comment on it's capacity.

Just occured to me that perhaps those of us who use this machine could
swap and share some of the analysis/emulation programs we have written.
Anyone out there interested?

In summary, if you're getting into SCSI development, something like this
can save your butt.  By the way, I have no ax to grind;  Just another
very satisfied customer!

 ___________________________________________________________________
|Regards,                           PPG Biomedical Systems          |
|Paul Matz                          One Campus Drive                |
|914-741-4685                       Pleasantville, NY. 10570        |
|{..}!uunet!philabs!ppgbms!paul     ppgbms!paul@philabs.philips.com |
|___________________________________________________________________|
path ppgbms!moe!paul@philabs.philips.com