[SunRescue] fun SCSI errors

[Greg A. Woods]

[ On Saturday, May 19, 2001 at 22:13:39 (-0500), Dan Debertin wrote: ]
> Subject: Re: [SunRescue] fun SCSI errors
>
> On Fri, 18 May 2001, Greg A. Woods wrote:
> 
> > How do you know that "termination is not a problem"?  Have you measured
> > the impedance of the bus and the termination power at both ends of the
> > bus?
> 
> Certainly not. I don't think I even know where to get tools for that...

Termination voltage (I should have used that more correct term! :-) is
usually relatively easy to measure if you can get at the drive near the
connector, and if you're careful, with just a voltmeter.  It is a lot
harder to do these days with high-density connectors, tiny drives, and
even more cramped chassis, etc.  Sometimes an adequate test is simply to
check if the light on an external active terminator is on.  Sometimes
they come on at the same brightness over a wide range of voltages, but
some seem to be more sensitive and thus useful for this purpose.  No
light is definitely a bad thing, but the presense of a "normal" light
unfortunately isn't always good enough.

As for impedance, well you can get tools to test at least cables, though
you wouldn't want them for home use (last one I looked at was ~$1500).
However I have encountered the odd store or corporate IT shop with one
that'll let me test my own cables on.  A really good store with a
service counter will usually offer to let you do this because then you
might buy a replacement cable if yours is bad.  :-)  [Of course you
should always demand to see the test done and verify the results with
your own eyes and against a good cable the of the stor shelf, just like
you should have the mechanic explain to you in detail why you need him
to expose the guts of your car engine for repair.]

> Yes; I actually was able to do this, on your suggestion, by plugging my
> NetBSD box directly into the SCSI shelf of the 670MP. Looked very odd ;).
> You were right; the NetBSD box reported problems as well.

I'm glad the test that showed the problem was that simple!

> I'm working with scroungeware -- whatever mix of disks, cables & adaptors
> I could pull together for the purpose. It looks like I might have a dodgy
> CEN50->CEN50 cable; swapping in one from work appears to have cleared
> things up.

I wasn't sure if you might not have enough similar parts on both buses
that you could swap until you found the bad one.

Anyway, glad you found the faulty part!

-- 
							Greg A. Woods

+1 416 218-0098      VE3TCP      <gwoods at acm.org>     <woods at robohack.ca>
Planix, Inc. <woods at planix.com>;   Secrets of the Weird <woods at weird.com>