HP3000-L Archives

May 1995, Week 4

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
"Rudderow, Evan" <[log in to unmask]>
Reply To:
Rudderow, Evan
Date:
Tue, 23 May 1995 18:33:00 EDT
Content-Type:
text/plain
Parts/Attachments:
text/plain (115 lines)
Jeff Vance <[log in to unmask]> wrote:
 
>Hi All,
 
<snip>
 
>...the intention of the Interex voters was to have easy command access.
 
I suspect that many of those who voted for a command intend to parse the
output of the command; that being the case, long file names become an issue.
 Both the name of the file being LISTF'd and the name of the program
accessing that file could conceivably be so long that the output line will
"wrap" to the next line; it seems to me that at that point any script
attempting to parse the output will head south.
 
Unless, of course, output directed to a file can exceed 80 character
records...But even is 255 character records are possible, a Posix file name
can be 1024 characters; so there's still a problem.  I think that rather
than right truncating the filename -- which cuts off the more "specific"
information -- I think that long filenames should be left truncated --
causing more "general" information to be dropped.
 
<snip>
 
Regarding Format 9:
 
>********************
>FILE: LOGFILE9.LOGGING1.SYSTEM12
>      NO ACCESSORS
>********************
 
How about "0 ACCESSORS" -- keeps the accessor count in a numeric motif. (And
by the way, what will the output look like when we get those weird CM KSAM
corner cases in which LISTF ,3 might report 0 accessors, 1 reader, 1
writer?)
 
<snip>
 
#S40     FRED,MGR.PAYROLL,PUB       -lock- REM: clientmpe.cup.hp.com
 PIN: 244      (/usr/local/bin/bsd/tools/some_program.exe)
 ACC: Read     REC#: 379         LOCKS: FLOCK-No,   OPEN-No,   GUFD-Wait
 
> - ",REMOTE" is displayed if any accessor is remotely connected to the
3000. If
>   all accessors are local then nothing extra is displayed.
 
Can you tell us who on the remote is accessing the file or otherwise give us
something so we can trace back to the origin?
 
>2) USER: line no longer shows ",GROUP" name...
 
That's funny, there seem to be ",GROUP" names in the user lines in the
examples...
 
> - For remote accessors the LDEV: label is replaced with REM: followed by
the
>   domain name for the incoming connection.  If the domain name cannot be
>   **quickly** obtained then the IP address is used instead.  If the domain
>   name is too long it wraps around.  If the process is not remote then
nothing
>   is shown.  A job streamed by a remote session is seen as a local job.
 
I'd rather have an IP address than a node name that's wrapped to tight (or
wrapped at all).
 
>3) The PIN: line has PIN followed by the program name starting in col 16.
 The
>   name is enclosed in parenthesis (like in SHOWPROC).  If the name is too
>   long it wraps to the next line.
 
Please, don't wrap.
 
<snip>
 
>New Choice 1  format 8 (file-open order still)
 
<snip>
 
>FILE: LOGFILE1.LOGGING1.SYSTEM12
>      5 ACCESSORS,SHARED,3 R,2 W,LOCKED,REMOTE
>#S12345  JOBNAME8,USER5678.ACCT5678 -lock- LDEV: 12345
>#S12345  JOBNAME8,USER5678.ACCT5678 -lock- LDEV: 12345
>#S40     FRED,MGR.PAYROLL,PUB       -lock- REM: clientmpe.cup.hp.com
>#J15     JSPOOL,RSPOOL.SYS,RSPOOL          SPID: O7654321
>#S12345  JOBNAME8,USER5678.ACCT5678 -lock- LDEV: 12345
 
In the example above, seeing #S12345 is redundant, and open count would be
more convenient.
 
<snip>
 
>------------------------------------------
>New Choice 2  format 8 (sort by j/s number)
>------------------------------------------
 
>FILE: LOGFILE1.LOGGING1.SYSTEM12
>      5 ACCESSORS,SHARED,3 R,2 W,LOCKED,REMOTE
>#S12345  JOBNAME8,USER5678.ACCT5678 -lock- LDEV: 12345
>#S40     FRED,MGR.PAYROLL,PUB       -lock- REM: clientmpe.cup.hp.com
>#J15     JSPOOL,RSPOOL.SYS,RSPOOL          SPID: O7654321
>
>Note: #S12345 only appears once now.
 
Much better, still a count of how many opens each session has would be nice.
 
BTW, I prefer choice 2 for both formats.
 
<snip>
 
>Thanks for all of your great input!
 
Thank you!
 
 -- Evan

ATOM RSS1 RSS2