Re: Unisys H/W failures

Paul Kimpel <paul.kimpel@xxxxxxxx> writes:
On 10/29/2009 2:20 PM, Louis Krupp wrote:
Chas wrote:
On Oct 29, 3:27 pm, Louis Krupp <lkrupp_nos...@xxxxxxxxxxxxxxxxx>
wrote:
Scott Lurndal wrote:
HVlems <hvl...@xxxxxxxxxx> writes:
On 27 okt, 18:15, sc...@xxxxxxxxxxxxx (Scott Lurndal) wrote:
215 was 47MB (180-byte sectors)
225 was 87MB (180-byte sectors)
The answer is as much confusing as it is detailed :-)
Confusing because my memory refuses to accept that 67MB disk packs
didn't exist, the value is 87 MB instaid!
However the 225 model sounds familiar, so it must have been an 87 MB
pack.
Well, with 100-byte sectors, the 225 is 65MB, which is close, but I
don't think Large systems (B[567]00) ever used 100-byte media.
If I recall correctly, all Burroughs disks and packs used 180-byte
segments.

Louis

I believe Medium Systems (2x00, 3x00, 4x00) used 100-byte segments.
At least, back when the 215s and 225s were used.

Interesting. I'm pretty sure Large Systems saw 180-byte segments on
those packs. Maybe they could be configured differently. My own
limited understanding was from a Large Systems software perspective.

Louis

Indeed, Large Systems and A Series only ever used disks with 180-byte
(30 word) segments. That was a format carried over from the B5500.

Medium Systems and V Series used disks with 100-byte sectors. The disk
pack media was the same for the two systems, but was formatted
differently. Anyone remember the IVR utility?

I think later in life the V Series systems were able to use media with
180-byte sectors, but seem to recall they still needed 100-byte units
for the MCP, and perhaps for object code storage.

100-byte media (known as DISK) was only needed for the MCP device. Codefiles
could reside on DISK (100-byte) or PACK (180-byte). Codefiles compiled
for DISK could be converted to PACK using the PCOPY utility.

A common disk subsystem in the 3500/4700 days was the 5N subsystem,
which utiltized head-per-track media providing 5ms seek times.

Later, 215, 225, 235, 206, 207, 677, 659 could be IVR'd as either DISK or
PACK. MD4, MD8, 680 and 682 could only be PACK (180-byte).

I/O controls and the first Host Transfer DLP's only supported decimal
addressing and medium systems used a 6-digit field for the disk address
limiting 207's to 1 millon sectors. Later controller firmware allowed
the 6 digit field to be treated as binary, allowing access to larger
media. V-series IOCB's make the address (C) field 8 digits.

http://vseries.lurndal.org/doku.php?id=peripherals:diskpacks
http://vseries.lurndal.org/doku.php?id=instructions:spio

scott
.

Relevant Pages

  • Re: PACK question
    ... disk for pack or for cursors. ... > rest of the disk with dummy files until the space left on the floppy was 0 ... > showed 3 temporary files and the table ...
    (microsoft.public.fox.helpwanted)
  • Re: VMS SIG discovers what Billy really relies on
    ... Imagine how the image would look with8 itanics in a rack with 3TB disk array ... head crash by mounting the crashed pack on another drive. ... installed into the first systems crashed drives into the second systems ...
    (comp.os.vms)
  • Re: Xah Lees Unixism
    ... > smD the TLA that represents a washing-machine size disk. ... pack was a little over 7mbytes. ... top & bottom row with 9 drives. ... i had written a report that relative disk system performance had ...
    (comp.lang.python)
  • Re: Xah Lees Unixism
    ... > smD the TLA that represents a washing-machine size disk. ... pack was a little over 7mbytes. ... top & bottom row with 9 drives. ... i had written a report that relative disk system performance had ...
    (comp.lang.lisp)
  • Re: PLEASE HELP!!! PACK AND GO ISN"T WORKING FOR ME!
    ... Extract it to your hard disk and then burn the files to a CD using your CD ... Or, to create an autorun CD, skip Pack and Go all together and check out ...
    (microsoft.public.powerpoint)