Re: speaking of filesystems [was: Definition of mainframe?]

2023-07-30 Thread Seymour J Metz
> A colleague said, "Everything is a file."  OS/360 aimed for that
> target with the abstraction of DD names, but missed by exposing
> hardware characteristics with "attributes" such as RECFM=FBM.

That was not an OS/360 design goal, but I don't believe that there is any 
conflict between attributes and "everything is a file" and, in fact, Unix has 
attributes. Where OS/360 falls short of that non-goal is that there is not a 
uniform naming system for "everything"; it's UNIT=foo instead of something like 
DSN=DEV.foo.

I would not classify RECFM as device dependent, but I would agree that, e.g., 
SPACE is, at least partially.


--
Shmuel (Seymour J.) Metz
http://mason.gmu.edu/~smetz3


From: IBM Mainframe Discussion List [IBM-MAIN@LISTSERV.UA.EDU] on behalf of 
Paul Gilmartin [042bfe9c879d-dmarc-requ...@listserv.ua.edu]
Sent: Saturday, July 29, 2023 8:50 PM
To: IBM-MAIN@LISTSERV.UA.EDU
Subject: Re: speaking of filesystems [was: Definition of mainframe?]

On Sat, 29 Jul 2023 18:26:29 -0400, Rick Troth wrote:
>
>Here's a neat trick: you can make a hard link to a sym-link.
>
I believe that's not required for POSIX conformance.  But I may
be misled by the lack of that ability in the "ln" utility.

>There are only a handful of actual file *types*:
>
>  * plain file
>  * directory
>  * character special (like a terminal)
>  * block special (like a disk drive or partition thereof)
>  * symbolic link
>  * and a couple others added by OMVS ... really
>
A colleague said, "Everything is a file."  OS/360 aimed for that
target with the abstraction of DD names, but missed by exposing
hardware characteristics with "attributes" such as RECFM=FBM.
OMVS recovered considerable with access methods that generate
RDWs, etc. as specified by the DCB.  (But they don't synthesize
the carriage control character. )

CMS doesn't even try, having distinct system calls for each device
type.

--
gil

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: speaking of filesystems [was: Definition of mainframe?]

2023-07-29 Thread Paul Gilmartin
On Sat, 29 Jul 2023 18:26:29 -0400, Rick Troth wrote:
>
>Here's a neat trick: you can make a hard link to a sym-link.
>
I believe that's not required for POSIX conformance.  But I may
be misled by the lack of that ability in the "ln" utility.

>There are only a handful of actual file *types*:
>
>  * plain file
>  * directory
>  * character special (like a terminal)
>  * block special (like a disk drive or partition thereof)
>  * symbolic link
>  * and a couple others added by OMVS ... really
>
A colleague said, "Everything is a file."  OS/360 aimed for that
target with the abstraction of DD names, but missed by exposing
hardware characteristics with "attributes" such as RECFM=FBM.
OMVS recovered considerable with access methods that generate
RDWs, etc. as specified by the DCB.  (But they don't synthesize
the carriage control character. )

CMS doesn't even try, having distinct system calls for each device
type.

-- 
gil

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


speaking of filesystems [was: Definition of mainframe?]

2023-07-29 Thread Rick Troth

I don't follow your comparison of PDS/e and Unix filesystems.
If I saw correlation of Linux filesystems with PDS, I glossed over it as 
stoopid. (Here again, I feel your pain.)


My understanding is that PDS is (historically) a means of segmenting one 
data set into related chunks. They're "related" because they're all 
members of the same data set.
The real "filesystem" for MVS is the catalog, and the "files" are the 
data sets, whether partitioned or not. Hopefully the VTOC and the 
catalog match-up. That's not always required.


PDS are also flat, unless something has changed recently.
PDS is more like the partition table on a PC disk. (Though the latter 
doesn't use names, per se, and tends to have less "members".)


You're absolutely right: a Unix filesystem is a file containing files. 
There the similarity to PDS ends.
Unix filesystems have two important features: inodes and directories. 
The inodes are usually invisible. The directories connect inodes with 
names, which is what people see.
A hard link is where two files (by name) refer to the same inode. A 
symbolic link is a special kind of Unix file that names another file. 
Here's a neat trick: you can make a hard link to a sym-link.

There are only a handful of actual file *types*:

 * plain file
 * directory
 * character special (like a terminal)
 * block special (like a disk drive or partition thereof)
 * symbolic link
 * and a couple others added by OMVS ... really


As I recall, other Unix vendors added their own special types. But I've 
slept since then and memory is fuzzy (or maybe I dreamed it).


Not all "filesystems" mountable on a Linux system properly implement 
inodes and directories. They therefore lack some functionality in 
practice; it can be significant.


Some features of Unix (and Linux) filesystems are a step up from 
historical MVS data sets. It is sadly still possible to have a data set 
with no time stamp.
So you can count that as one thing Unix did that MVS has learned 
afterward. (Not a slam on z/OS. Just keeping things in perspective.)


-- R; <><


On 7/29/23 12:28, Jon Perryman wrote:

Can anyone provide the definition of MAINFRAME? The ARS Technica article is 
complete nonsense because the mainframe is a state of mind and nothing to do 
with reality. Can anyone prove me 
wrong?https://arstechnica.com/information-technology/2023/07/the-ibm-mainframe-how-it-runs-and-why-it-survives/.

The IBM z16 is just 4 motherboards containing 16 CPU and many PCIe slots. Linux 
will run on an IBM z16. Is a PC also mainframe? Forget zPDT because I suspect 
it still uses a PCIe zCPU card. I can't say with any certainty, but I suspect 
that z/OS will run on a PC by using Hercules. What is the definition of 
MAINFRAME?

1. CPU does not make a mainframe: The smallest IBM z16 (39 user cores of the 64 
cores) is the same as an AMD Ryzen 4.2Ghz CPU (64 user cores of 64 cores). The 
largest IBM z16 (200 user cores of the 256 cores) is the same as 4 AMD Ryzen 
CPU on 1 motherboard (256 user cores of the 256 cores). Both are CISC CPU (AMD 
uses X86 instructions versus IBM z instructions). IBM Telum (5.2Ghz) has a 
slightly faster clock than AMD Ryzen (4.2Ghz) but is offset by the 25% extra 
user cores. IBM z16 has 4 motherboards for 16 CPU and the same AMD Ryzen would 
need 1 motherboard for 4 CPU.

2. Hardware does not make a mainframe. IBM z16 has PCIe and ram which are also 
on every modern motherboard. IBM z16 chooses not to include other hardware 
(e.g. SATA, IDE, WIFI and more). Motherboards choose not to have 1,600 PCIe 
slots. IBM could allow PCIe graphics cards, mice, keyboards and more. 
Essentially, IBM z16 and AMD Ryzen can implement the same hardware if there was 
enough customer demand.

3. OS does not make a mainframe. Linux running on z16 doesn't make it mainframe 
Linux. There's nothing stopping Linux from taking advantage of every z16 
hardware feature (e.g. 1,600 PCIe slots) but no one is willing to build the 
Linux software. IBM hasn't duplicated z/OS software features in Linux.

4. Software does not make a mainframe. IBM sells DB2 for Linux and DB2 for 
z/OS. DB2 for Linux runs on all hardware including z16. With Linux, you can 
still run DB2 on z16 but large customers choose DB2 for z/OS.

ASK YOURSELF: Other than design philosophy, name 1 fundamental difference 
between IBM z16, AMD Ryzen and the software.

ASK YOURSELF: Since design philosophy is the only difference, name the 
philosophy that makes a mainframe.

Despite the story's false claims for z/OS relevance, it is ignorance in the 
Linux community that makes IBM z/OS relevant. Specifically, it's the lack of 
design in Linux. Consider DB2 for Linux and DB2 for z/OS which are the same 
product both from IBM and available on an IBM z16. Linux people tell you they 
provide the same results, but they ignore the intrinsic capabilities of z/OS 
design. DB2 for Linux supports high availability and large databases but it 
requires knowledge of big data solutions, Linux