[566] in linux-scsi channel archive
Re: Errors & Unit Attention
daemon@ATHENA.MIT.EDU (Stephen D. Williams)
Wed Aug 30 20:06:42 1995
From: sdw@lig.net (Stephen D. Williams)
To: linux-vger@wab-tis.rabobank.nl
Date: Wed, 30 Aug 1995 10:25:10 -0400 (EDT)
Cc: sdw@lig.net, bsa@kf8nh.wariat.org, eric@aib.com, lnz@dandelion.com,
linux-scsi@vger.rutgers.edu
In-Reply-To: <HAA14590@sys3.pe1chl.ampr.org> from "Rob Janssen reading Linux mailinglist" at Aug 29, 95 07:23:14 am
>
> According to Stephen D. Williams:
> > > Try it with a PC SCSI device sometime. AHA1542s in particular will freeze
> > > solid - the on-board microcode crashes and you have to reset/powerdown to
> > > get it back. (I have had it happen.) Most other PC SCSI cards of my
> >
> > This is one of the reasons that I will NEVER use Adaptec again.
> > I've had it do this, and burn out the SCSI fuse, with attempted use
> > of 8mm tape drives...
> >
> > I've had so much better results with Future Domain it's not even a contest...
> >
> > (Note: most of my SCSI hacking/assembling days was in 8-bit/16-bit era...)
>
> I have no complete overview of the Future Domain product line, but what
> I have seen of it was in the lower-functionality/performance range.
> (more like the AHA1522)
No, just their really cheap boards. They also have made the chips used
by other OEM boards. (Seagate st01/2...)
They also have VL/PCI LB, EISA, differential, fast/wide, etc.
I'm sure the higher end boards have their own microcode, etc.
I think a huge difference in reliability was mainly that I believe
that all FD's use a type of active termination (which is so good I've
even run without terminators) whereas the 1542a-c (I've fought with
all three at various customers) had very finicky passive termination.
I gave up ever getting more than 3 drives to work with Adaptec, but
did it ages ago even with a FD950 (a $60 8-bit card).
I currently use FD1680's for my workhorse system... (Which is now
outdated.)
> Adaptec also makes high-end adapters with busmastering and a processor
> that executes a program on its own (called microcode in the quoted text).
> Of course, when you don't have that it can't crash either.
>
> Blown fuses are normally caused by misaligned connectors, not by the
> host adapter itself.
I don't think so in this case. The 8mm may have had a software problem,
but after a while of successful data transfer, the 8mm activity light
went solid and the fuse was blown. No problem with FD.
> Rob
>
> --
> +------------------------------------+--------------------------------------+
> | Rob Janssen rob@knoware.nl | AMPRnet: rob@pe1chl.ampr.org |
> | e-mail: pe1chl@wab-tis.rabobank.nl | AX.25 BBS: PE1CHL@PI8WNO.#UTR.NLD.EU |
> +------------------------------------+--------------------------------------+
sdw
--
Stephen D. Williams 25Feb1965 VW,OH (FBI ID) sdw@lig.net http://www.lig.net/sdw
Consultant, Vienna,VA Mar95- 703-918-1491W 43392 Wayside Cir.,Ashburn, VA 22011
OO/Unix/Comm/NN ICBM/GPS: 39 02 37N, 77 29 16W home, 38 54 04N, 77 15 56W
Pres.: Concinnous Consulting,Inc.;SDW Systems;Local Internet Gateway Co.;28May95
