🚀 go-pugleaf

Sure, it's worth looking into.  Sequent charges about $20,000 for a Fujitsu
M2344K (550 Mb formatted) disk drive, with mounting rack and power supply.  I
get these disks from a third party supplier for $8,000, complete.

The disks from Sequent come formatted and tested, and are easier to install,
but we've saved $12,000 on each of the eight drives I've installed, which I
feel is worth the extra time and hassle involved.

Installation the drives and cables takes about 1.5 hours, formatting takes
between 3 and 5 hours per drive, depending on the total number of errors.  The
formatting procedure from Sequent's DYNIX Administration manual is correct and
complete.

If anyone is interested in more detailed information, I will mail or post my
written procedure for purchasing and installing M2344K's on a Sequent system.

Patrick Wolfe
System Manager, Kuck and Associates, Inc.
Patrick.Wolfe@kailand.kai.com  -or-  kailand!Patrick.Wolfe

In article <1049@orion.cf.uci.edu>, roode@orion.cf.uci.edu (Dana Roode) writes:
> We need to add some additional disc to our Sequent Symmetry system,
> and are wondering if it makes sense to look into third party
> disc.  Does anyone at other Sequent sites have specific experience
> in using non-Sequent disc?
> 
>      Dana Roode
>      UC Irvine Computing Facility
> 
> -- 
> Dana Roode

We at CLSI (a Sequent OEM) have added several of the CDC disks:

		9715 - 340 MB (FSD)
		9771 - 800 MB (XMD)
		9720 - 368 MB and 850 MB (EMD)

We will soon be working on the 9720 series 1.2 Gig drive.

These drives work very well on the DCC (firmware version 12 or later).

BUT!!!!  The DCC is a VERY intelligent controller.  So intelligent, you have
to be intimately familiar with it and the drive to get the timing right.

We had to take a three-day course from Sequent to figure this whole thing
out.  Let's just say that adding drives is a non-trivial project.

Of course, theoretically, you could add any SMD-interface drive to the DCC
if you had enough information.

Sequent supports many of the Fujitsu drives.  These drives are (among) the
best on the market (in terms of reliability).  [forgive the editorializing
here; I've been a fan of Fuji for some time].

I hope these idle wanderings of the mind help.  If I can help you with
specific questions, send me some mail.

			Peter Craine
			 ..!harvard!bbn!clsib21!peter
			   [bah! who needs a .signature]

In article <1049@orion.cf.uci.edu> roode@orion.cf.uci.edu (Dana Roode) writes:
>We need to add some additional disc to our Sequent Symmetry system,
>and are wondering if it makes sense to look into third party
>disc.  Does anyone at other Sequent sites have specific experience
>in using non-Sequent disc?
>
>     Dana Roode
>     UC Irvine Computing Facility
>
>-- 
>Dana Roode

We use a lot of System Industries equipment on our Symmetry S81 
(upgraded from a Balance 21k).  As far as disk drives go, we have
used the SI83, SI93, Swallows and Eagles from SI with no problems.

The SI83 is a 690 megabyte (unformatted) dual density Swallow that 
fits two to a tray 1/2 the height of an Eagle (so 4 SI83s will fit 
in the footprint of an Eagle).  The SI93 is actually a NEC D2363 
(1.132 GB unformatted).  Two of these will fit in the footprint of 
an Eagle.  Access times for the SI83 and SI93 are 16 ms and 15 ms 
respectively.  The SI Swallows and Eagles are identical to what 
Sequent supplies.

We had to update the partition table of the zd disk driver (3.0.4 
on up) to make it work with the NEC drive (no source needed).  
The SI83 will work with the stock 3.0.4 on up.  

Right now we have 4 SI83s, 8 SI93s, 1 Sequent Swallow and 1 SI Swallow
(as a backup for the one Sequent disk), about 11 gigabytes of disk space.
The only drawbacks are that we have yet to figure out how to run Sequent's 
disk diagnostics on the SI83 or SI93 and it takes 1/2 the night to format
an NEC drive.

We've also interfaced the SI STC tri-density 125 ips 9-track tape drive 
and SI line printer (a Fujitsu M3043) on the Symmetry.  If anyone is 
interested in any of this, drop me a line.

I'd also like to know what other 3rd party equipment people are using on 
their Symmetry or Balance.

Richard Loo 			UUCP:     ..!{oberon,elroy}!hacgate!rich
(213) 648-0239			Internet: rich@hacgate.scg.hac.com

Hughes Aircraft Company
PO Box 92919 S65/J305
Los Angeles, Ca 90009

/* Written  1:01 pm  Oct  3, 1988 by peter@clsib21.UUCP in kailand:comp.sys.sequent */
>
> BUT!!!!  The DCC is a VERY intelligent controller.  So intelligent, you have
> to be intimately familiar with it and the drive to get the timing right.
>
> We had to take a three-day course from Sequent to figure this whole thing
> out.  Let's just say that adding drives is a non-trivial project.
>
> Of course, theoretically, you could add any SMD-interface drive to the DCC
> if you had enough information.

Adding disk drives is ONLY non-trivial if you try to add drives that Sequent
doesn't already support.  A quick peek into /etc/disktab shows which drives
they currently support on the DCC:  Fuji M2333K, M2351A (eagle), M2344K,
M2382K, CDC 9715-340, and 9771-800.  Adding those drives is relatively
trivial, compared to adding unsupported drives.

One problem (detailed in a talk given by Purdue University at the 1987
Sequent User's Group meeting) is that the disk format programs Sequent
delivers has the drive parameters hard coded in, instead of reading them from
a user-editable table.  So even if you know all the information that the format
program needs, you need the *source* code to be able to add drives.  This
eliminated any chance we had for adding non-supported drives.

Patrick Wolfe
System Manager, Kuck and Associates, Inc.
Patrick.Wolfe@kailand.kai.com  -or-  kailand!Patrick.Wolfe

In article <919@clsib21.UUCP> peter@clsib21.UUCP (Peter Craine) writes:
>Sequent supports many of the Fujitsu drives.  These drives are (among) the
>best on the market (in terms of reliability).  [forgive the editorializing
>here; I've been a fan of Fuji for some time].

Ahem!  Fuji makes film, not disk drives.


			Ron
-- 

      Ronald O. Christian (Fujitsu America Inc., San Jose, Calif.)
      {amdahl, pyramid, sun, unisoft, uunet}!fai!ronc -or- ronc@fai.com

      Calling all Fujitsu Usenet sites!  Contact fai!ronc or
      ronc@fai.com to establish uucp connection.

In article <2400024@kailand> pwolfe@kailand.KAI.COM writes:
>
>One problem (detailed in a talk given by Purdue University at the 1987
>Sequent User's Group meeting) is that the disk format programs Sequent
>delivers has the drive parameters hard coded in, instead of reading them from
>a user-editable table.  So even if you know all the information that the format
>program needs, you need the *source* code to be able to add drives.  This
>eliminated any chance we had for adding non-supported drives.

This is no longer true.  In the *binary* distribution of the DYNIX V3.0,
the proper files are available.  If you look in /usr/sys/stand/conf_zdc.c,
you will find the table that we used to have to change by only having the
source code available.  In fact, Sequent now has a Makefile available in
/usr/sys/stand to remake the stand-alone programs (like the formatter --
zdformat -- if you plan to boot off of the device, also remake "boot" and
"zdboot" -- there are other caveats if you are interested).  In conf_zdc.c,
you need to come up with the magic configuration numbers that are described
in the zdinfo structure (described in /usr/sys/stand/zdc.h -- but the most
important chunk -- zdcdd -- is described in /usr/sys/zdc/zdc.h).  One of
the Sequent systems (this happened to be a Symmetry) at Purdue (the
Department of Computer Science) was able to add support for a CDC 9766
essentially following the directions that Mike Marsh and I laid out in the
paper with only the "binary" distribution (not source code) with DYNIX
V3.0, so I know it is now possible.  Thanks to Sequent for adding this
support!

Kevin C. Smallwood			Internet: kcs@j.cc.purdue.edu
Manager of UNIX Systems			UUCP: ...!{pur-ee,purdue}!kcs
Purdue University Computing Center

> 
> One problem (detailed in a talk given by Purdue University at the 1987
> Sequent User's Group meeting) is that the disk format programs Sequent
> delivers has the drive parameters hard coded in, instead of reading them from
> a user-editable table.  So even if you know all the information that the format
> program needs, you need the *source* code to be able to add drives.  This
> eliminated any chance we had for adding non-supported drives.
> 

Yes, you need some source, but the only source you need is in the
distribution kits.  There are five files to modify:
		/sys/stand/conf_zd.c
		/sys/zdc/zdc.h
		/sys/conf/conf_zd.c
		/sys/conf/controller.balance (only potentially)
		/etc/disktab

The real problem is not having the source, but understanding what must be
changed/added to support the drives.  In particular, /sys/stand/conf_zd.c
is the "non-trivial" portion of adding support I alluded to in my
previous posting.

So, you don't need any more source, but this doesn't make the problem any
less severe.

				Peter Craine

When the Sequent User Group met in 1987, the current release of DYNIX was V2.1.
I was not aware that things had changed with V3.0.

Thanks you Sequent for making this code available!

Patrick Wolfe
System Manager, Kuck and Associates, Inc.
Patrick.Wolfe@kailand.kai.com  -or-  kailand!Patrick.Wolfe