360/370 disk drives 
Author Message
 360/370 disk drives



Quote:
>Ah yes.  A job that did a lot of seeks on a 2311 would cause the
>whole unit to vibrate like an out-of-balance washing machine.  I used
>to work with a girl who loved to sit on top of the drives when they
>were seeking wildly.

I once had a boss who was considering 3330 compatible drives from
Memorex. He wanted to compare the drives and asked me to write a
program that would seek to the center, seek back one track, seek
forward two tracks, . . ., until it had visited every cylinder. I told
him that the program was trivial but that it would give no timing
information relevant to real loads. He told me to write it anyway.

I called it DANCE, because that's what it made the drives do.

--
     Shmuel (Seymour J.) Metz, SysProg and JOAT

Any unsolicited commercial junk E-mail will be subject to legal
action.  I reserve the right to publicly post or ridicule any
abusive E-mail.

I mangled my E-mail address to foil automated spammers; reply to
domain Patriot dot net user shmuel+news to contact me.  Do not



Fri, 15 Jul 2005 00:59:23 GMT  
 360/370 disk drives


Quote:
>IIRC (war story from guy at PPOE) VTOC entries were 80 byte blocks,
>so could be rebuilt from listings using card reader.

No. Labels were 80 bytes; the DSCB was substantially larger.

--
     Shmuel (Seymour J.) Metz, SysProg and JOAT

Any unsolicited commercial junk E-mail will be subject to legal
action.  I reserve the right to publicly post or ridicule any
abusive E-mail.

I mangled my E-mail address to foil automated spammers; reply to
domain Patriot dot net user shmuel+news to contact me.  Do not



Fri, 15 Jul 2005 00:49:51 GMT  
 360/370 disk drives


Quote:
>By the way, just before I left IBM, I understand that there was a new
>type of dataset in MVS called a PDSE (like PDS "Enhanced" or some
>such). I never got a chance to play with them because the project I
>was working on got cancelled, and the system programmer hadn't
>installed the necessary software to make PDSE's work.

it came in with DF/SMS, but load modules (program objects) were not
supported until R2. Fixed length blocks under the covers.

Quote:
>So I don't know if they made PDSE's compatible with the old-style
>PDS,

As long as you stuck to BPAM, BSAM and QSAM. Programs using EXCP can't
handle it, and the low-level interfaces are not documented for
customers.

--
     Shmuel (Seymour J.) Metz, SysProg and JOAT

Any unsolicited commercial junk E-mail will be subject to legal
action.  I reserve the right to publicly post or ridicule any
abusive E-mail.

I mangled my E-mail address to foil automated spammers; reply to
domain Patriot dot net user shmuel+news to contact me.  Do not



Thu, 14 Jul 2005 23:52:03 GMT  
 360/370 disk drives
On Sun, 26 Jan 2003 11:59:23 -0500, Shmuel (Seymour J.) Metz hath writ:

Quote:


>>Ah yes.  A job that did a lot of seeks on a 2311 would cause the
>>whole unit to vibrate like an out-of-balance washing machine.  I used
>>to work with a girl who loved to sit on top of the drives when they
>>were seeking wildly.

> I once had a boss who was considering 3330 compatible drives from
> Memorex. He wanted to compare the drives and asked me to write a
> program that would seek to the center, seek back one track, seek
> forward two tracks, . . ., until it had visited every cylinder. I told
> him that the program was trivial but that it would give no timing
> information relevant to real loads. He told me to write it anyway.
> I called it DANCE, because that's what it made the drives do.

At a PPOE we 'tried out' the Itel 3330-look-alikes -- on a (yes)
Sys/360 Mod65  --  back about circa 1971.  (Itel had a Magic Box
Selector Channel interface adapter, and a wad of source changes to IOS.
[Thread drift: remember when SYS/360 was `Open Source'?])  
Long story shorter:  One day the operators called in to the
Systems Programming Cube Farm to report that the system was  
slowing down.  Smug know-it-alls that we were: "Ya, sure."
I went in and walked up to the 1052 and entered "DA".  It took
about 5 seconds before the first line of reply appeared!
Hmmmm!!
I scratched my head -- watched the blinken lights on the CPU
console -- then I scratched somewhere else -- and,
unbelievingly, I tried DA again.
This time it took 9 seconds for the first line appeared!!
I turned around and absently stared out into the computer room --
trying to construct some scenario on how this could be.
My diffused focus fell on the string of Itel drives (they
were single cabinet style -- ala the old 2311's.) and I noticed
that I could  a l m o s t  read the manufacture's label on
SYSIPL.    !!!!!!!???????   The drive used a big wide, motor-driven
belt to power the disk spindle.  And, sure enough, opening
the front 'modesty' panel on the drive revealed that the belt
had been thrown.  When I first saw it spinning down, SYSIPL was
running at about 60-100 RPM.  It was running at about 30 RPM
by the time we managed to semi-gracefully get everything shut
down.  Yes, we knew that, from about 2000 RPM and on down, the
R/W heads were in full contact with the surface of the platters.
It was the `softest` head crash I ever saw.

Among other things, the SVC transient area(s) sure took
a Bit Hit in performance.

Jonesy
--
  | Marvin L Jones       | jonz           |  W3DHJ   |  OS/2

  |   7,703' -- 2,345m   |   frontier.net |  DM68mn             SK



Fri, 15 Jul 2005 02:56:10 GMT  
 360/370 disk drives


Quote:
>No.  If all you're doing is an 80-80 list (or anything similar, for
>that matter) you need multiple cycles only if you need to go through
>the card's data multiple times.

You are, of course, correct: I was thinking of the 716 printer on the
7090 rather than the 407.

--
     Shmuel (Seymour J.) Metz, SysProg and JOAT

Any unsolicited commercial junk E-mail will be subject to legal
action.  I reserve the right to publicly post or ridicule any
abusive E-mail.

I mangled my E-mail address to foil automated spammers; reply to
domain Patriot dot net user shmuel+news to contact me.  Do not



Fri, 15 Jul 2005 02:55:30 GMT  
 Card Columns (was Why did they make ... ?)


Quote:
>You could read only 72 of the 80 columns.  A different board could
>choose a different 72 columns, but the norm was 1-72.

You're right; I was confusing the 711 with the 716.

--
     Shmuel (Seymour J.) Metz, SysProg and JOAT

Any unsolicited commercial junk E-mail will be subject to legal
action.  I reserve the right to publicly post or ridicule any
abusive E-mail.

I mangled my E-mail address to foil automated spammers; reply to
domain Patriot dot net user shmuel+news to contact me.  Do not



Fri, 15 Jul 2005 02:53:08 GMT  
 Card Columns (was Why did they make ... ?)

Quote:



> >You clipped that line.   The VALUE OF ID "XXXYYZ" identified the the
> >file.   No JCL needed.

> No, what is sounds like is that the JCL was built into the program and
> you had to recompile every time it changed.

IIRC, there *was* a control card you could use to override (i.e.JCL),
but I don't believe it was much used.


Fri, 15 Jul 2005 03:56:43 GMT  
 Card Columns (was Why did they make ... ?)
On Sun, 26 Jan 2003 12:04:33 -0500, "Shmuel (Seymour J.) Metz"

Quote:



>>You clipped that line.   The VALUE OF ID "XXXYYZ" identified the the
>>file.   No JCL needed.

>No, what is sounds like is that the JCL was built into the program and
>you had to recompile every time it changed.

Other OSes have these things called defaults and file systems,
rather than access method libraries, so no need to give the
program a list of access method parameters every time a file is
accessed.

Thanks. Take care, Brian Inglis         Calgary, Alberta, Canada
--

    fake address                use address above to reply






Fri, 15 Jul 2005 04:04:32 GMT  
 Card Columns (was Why did they make ... ?)

Quote:

> Other OSes have these things called defaults and file systems,
> rather than access method libraries, so no need to give the
> program a list of access method parameters every time a file is
> accessed.

Nor need you in OS/360 and its many follow-ons.  Assuming the file
exists and is cataloged, you need only give:

//ddname DD DSNAME=filename,DISP=SHR|OLD|MOD

If you know you're really hammering a QSAM file, DCB=BUFNO=nn can be
added to boost performance, though the effect is less than it used to be
(because the default is has long been 5 instead of the original 2).

When creating a new file, all that is generally needed is:

//ddname DD DSNAME=name,DISP=(NEW,CATLG),
// UNIT=something,SPACE=(blocksize,(primary,secondary),RLSE),
// DCB=BLKSIZE=blocksize

Utility programs may need:

// DCB=(RECFM=format,LRECL=recordsize,BLKSIZE=blocksize)

since they, by definition, are written to be flexible, but
most programs will have RECFM and LRECL hard-coded.

The complexity of JCL has been grossly exaggerated by the tendency of
textbooks to do an infodump.  The design wasn't perfect when it was
made, and has been rendered rather old-fashioned by much new technology,
but it's nowhere near as {*filter*}y-awful as many people think.

--
John W. Kennedy
"The poor have sometimes objected to being governed badly;
the rich have always objected to being governed at all."
   -- G. K. Chesterton, "The Man Who Was Thursday"



Fri, 15 Jul 2005 12:38:03 GMT  
 Card Columns (was Why did they make ... ?)
On Mon, 27 Jan 2003 04:38:03 GMT, "John W. Kennedy"

Quote:


>> Other OSes have these things called defaults and file systems,
>> rather than access method libraries, so no need to give the
>> program a list of access method parameters every time a file is
>> accessed.

>Nor need you in OS/360 and its many follow-ons.  Assuming the file
>exists and is cataloged, you need only give:

>//ddname DD DSNAME=filename,DISP=SHR|OLD|MOD

>If you know you're really hammering a QSAM file, DCB=BUFNO=nn can be
>added to boost performance, though the effect is less than it used to be
>(because the default is has long been 5 instead of the original 2).

>When creating a new file, all that is generally needed is:

>//ddname DD DSNAME=name,DISP=(NEW,CATLG),
>// UNIT=something,SPACE=(blocksize,(primary,secondary),RLSE),
>// DCB=BLKSIZE=blocksize

>Utility programs may need:

>// DCB=(RECFM=format,LRECL=recordsize,BLKSIZE=blocksize)

>since they, by definition, are written to be flexible, but
>most programs will have RECFM and LRECL hard-coded.

>The complexity of JCL has been grossly exaggerated by the tendency of
>textbooks to do an infodump.  The design wasn't perfect when it was
>made, and has been rendered rather old-fashioned by much new technology,
>but it's nowhere near as {*filter*}y-awful as many people think.

It's the knowledge required to fill in the blanks that is the
barrier to understanding. Most programmers have a hard time
picking reasonable values -- they just don't want to know about
the details -- an order of magnitude or two plus or minus is
close enough for them, or whatever the analyst estimates. System
programmers always had to QA JCL and change the values
appropriately in the shops I worked at.

Thanks. Take care, Brian Inglis         Calgary, Alberta, Canada
--

    fake address                use address above to reply






Fri, 15 Jul 2005 16:42:46 GMT  
 Card Columns (was Why did they make ... ?)

Quote:


>>No.  If all you're doing is an 80-80 list (or anything similar, for
>>that matter) you need multiple cycles only if you need to go through
>>the card's data multiple times.
>You are, of course, correct: I was thinking of the 716 printer on the
>7090 rather than the 407.

I never had to actually write code to use the 716, but I seem to recall
that in addition to requiring additional print cycles for more than
72 characters (more exactly, more than 24 words of data) you would
also need an I/O event to send appropriate signals to the hubs to
suppress the linefeed.  Or do I have it backwards, and the carriage
didn't move until you sent an order for a do-the-linefeed pulse
to be emitted?

I think I'm going to have to go on an archaeological expedition to find
my 7090 PoP where this is documented.

Joe Morris



Fri, 15 Jul 2005 21:59:15 GMT  
 
 [ 373 post ]  Go to page: [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25]

 Relevant Pages 
 

 
Powered by phpBB® Forum Software