Hidden Value May 1999

Hidden Value details commands and procedures in MPE (and some in Vesoft’s MPEX) that can improve your productivity with HP 3000 systems. Get a free NewsWire HP 3000 Always Online cap — submit your MPE tip directly to us here at the NewsWire. Send your tips to editor@3000newswire.com, or fax them to 512.331.3807.

Edited by John Burke

Bug or Feature? Mark Bixby, who also provided a workaround, asked this question:
I just discovered that :LISTFILE isn’t computing pattern length correctly:
:listfile [A-Z][A-Z][A-Z]????[A-Z],2
File name is more than eight characters long. (CIERR 532)

Bzzzt! Wrong! That pattern is indeed an eight-character standard MPE file name. But, for an easy workaround you can use HFS syntax:

:listfile ../[A-Z][A-Z][A-Z]????[A-Z],2
I haven’t checked the other commands that support [A-Z] pattern syntax, but it would not surprise me if they had this same problem.

Jeff Kell adds:
In the event that some engineer should happen to look at fixing this, it also generates this error for the ‘@’ wildcard (and the ‘@’ may not be present). I’ve seen this error while looking for source code or ancillary files to a program file, e.g.,

:listf @PGMNAME@.@,2 (looking for source to PGMNAME) File name is more than eight characters long. (CIERR 532)
How do you change the TCP port the Apache/iX server listens to? I have tried some changes in the httpd.conf file, but I do not notice any change in the port behavior.

Andreas Schmidt replies:
/APACHE/PUB/conf/httpd.conf is the right file!
Here you have the entry PORT 80, which can be changed. Restart the server job and the new port is used. But pay attention to the fact that the system must know the new port! (NMMGR)

Is there any way to easily generate random passwords on the HP 3000?

Chris Bartram replies:
Pick up a copy of rndpass from www.3kassociates.com in the public domain software area. It accepts a list of account, user, and/or group names and assigns them each random passwords. I’ve been using it for many years to randomize passwords on unused accounts/groups/users.

Donna Garverick further contributed:
Robelle wrote up something about how to use Suprtool and Robelle’s (spelling) dictionary to generate ‘random’ passwords. It should be available at their Web site . It looks really interesting. I think I used rndpass awhile back when I was told (at 4 in the afternoon, of course) to change all the account passwords on one of my systems. Boy, you should have seen me trying to explain what kind of word ‘XUKEROZ’ is. I think Robelle’s solution is more acceptable to most users (assuming of course you have Suprtool).

We received some software (COBOL subroutines) in an RL from a third party. Is there any way I can determine if it is compatibility mode or native mode? LINKEDIT doesn’t give this information; however, I do notice the compatibility mode switches indicator in GLANCE when it’s running.

Stan Sieler replies:
If the filecode (reported by LISTF, 2) is RL (or PROG or USL or SL) then the code is CM (or, at best, Object Code Translated — which is still CM, albeit somewhat faster). If the filecode is NMRL, NMXL, NMOBJ, NMPRG, then the code is NM.

Can Communicator 3000 articles be found on the Web?

In general, documentation can be found at www.docs.hp.com and in the specific case of 3000 Communicators, at www.docs.hp.com/mpeix/#com3000.

How do you get back to factory defaults with an HP JetDirect Ex? I cannot telnet to it. I can ping it but JetAdmin cannot see it.

Gilles Schipper replies:
You can reset it to factory defaults as follows (see page 6.6 of the JetDirect Ex Plus manual):

1. Unplug the power cord.
2. While holding down the test button (which is the purple-colored button between the 2 lights), plug the power cord back in and continue to depress the test button for five seconds more.

I noticed prior to doing a shutdown that the LED on LDEV 1 was almost constantly lit. I had already shut down all network services and aborted all jobs and sessions. This continued for about 15 minutes before I rebooted. Using SOS from Lund, the drive activity showed about 50 percent utilization and 30 I/Os per second. We have 25 other drives that show utilization at less than five percent and I/Os less than five per second. We have only system files on LDEV 1. HP says it is normal for LDEV 1 to get hit hard by I/O activity. Is this the norm for everyone else, or do we have a problem with LDEV 1?

Bill Lancaster replies:
There are two DTC-related events that are now logged by default that are creating all these I/Os. If you look around in NMMGR for the DTC configuration, you’ll see a place where you can configure six different logging events. The default used to be YNNNNN but is now YYYNNN. Change it back to YNNNNN, do a DOIONOW, and you’ll be all set.

I have a “saved” a spool file (it is in the ;SPSAVE format). I want to now move that file back to the READY state, but I can’t find out how. HELP doesn’t really help.

Andreas Schmidt replies:
SPOOLF ;SELEQ=...;PRINT should do the job.

Harlan Lassiter and Jeff Woods both noted:
The command works by COPYing the data to a new spool file rather than simply relinking the existing spool file. The attributes of the old spool file are used for the new copy unless specified otherwise on the SPOOLF command.

I need to have a batch job wait for n seconds, but I’d also like it to “wake up” immediately if needed. I was thinking that a message file would do the trick.

[The poser of the question went on to describe what he had tried to do with message files. Several people contributed their experiences in trying to make such a scheme work or ideas on how it might work. But problems remained. Robert Schlosser came to the rescue by noting a new feature in MPE/iX 6.0:]

You did not specify which version of MPE you are running. If you are running on 6.0, try using the new flavor of the PAUSE command.

PAUSE 60;JOB=wakeup;NOTEXIST
The job will pause there for 60 seconds or until a job named WAKEUP shows up on the system. See if this does it for you.

We are getting ready to install PowerPatch 6 for MPE/iX Release 5.5. How can we discover what patches have already been installed? Or whether a particular patch has been installed?

Jan Gerrit Kootstra and Andreas Schmidt reply:
You can give the command print hpswinfo.pub.sys and look for the appropriate patches. Another way to find a specific patch is:

run sh.hpbin.sys cat /SYS/PUB/HPSWINFO | grep ... exit
Fill in the patch name in capitals after grep.

How do you determine the PIN of the currently running program programmatically? I’ve looked at PROCINFO, WHO etc. in the intrinsic reference manual, but drew a blank.

Ken Vickers replies:
Look again. Supply a PIN of zero to PROCINFO and it will return info about the calling process:

PROCINFO(&iStat1,&iStat2,0,1,&MyPin);
Mike Hornsby concurred, and also suggested:
A simple method would be to use the command intrinsic with a SHOWPROC piped to a file. Your program could then parse the file and get not only the PIN but also the tree structure and CPU information. Note: In the following example I did this from within EDITOR to show what a SHOWPROC from within a program would look like.

/:showproc >purgeme /:print purgeme QPRI CPUTIME STATE JOBNUM PIN (PROGRAM) STEP C152 0:02.065 WAIT S151 61 :EDITOR C152 0:00.109 READY S151 70 (EDITOR.PUB.SYS)
HP’s Jeff Vance noted:
There is a CI variable named HPPIN that contains the PIN of “you”.

We are running MPE/iX 5.5 on a 987. We had a job abort when it went into KSAMUTIL and attempted a KEYINFO on an alleged KSAM file. The error was “THE REFERENCED FILE IS NOT A KSAM FILE (1062)” So, I did the natural thing and did a LISTF,2 and it appeared to be right!
FILENAME CODE ------------LOGICAL RECORD----------- ----SPACE---- SIZE TYP EOF LIMIT R/B SECTORS #X MX ARDBFKD * 70B FAK 257744 465513 53 158880 78 * So I went into MPEX and did the same thing and got the following results: FILENAME CODE ------------LOGICAL RECORD----------- ----SPACE---- --DAYS-- SIZE TYP EOF LIMIT R/B SECTORS #X MX ARDBFKD * KSAM 70B FAK 257744 465513 53 158880 78 * current 1041442 3 readers

The FILE TYPE shows a K and I couldn’t find anything in the on-line help to see what it meant. What is happening here?

Tom Renz and Gavin Scott reply:
What you have is a Native Mode (NM) KSAM file instead of a Compatibility Mode (CM) KSAM file. NM KSAM files are indicated by the “K” in the TYP column in the LISTF command. It’s probably a bug/feature of MPEX that shows the file as a “KSAM” file in the CODE column. This is because the file code is actually a “0” (zero) when it is built, unless you specify differently when using the BUILD command or by altering it in MPEX.

Regarding KSAMUTIL, this utility only works with CM KSAM files and does not know anything about NM KSAM files. With NM KSAM files you can use “LISTF,7” to obtain the keys and their definitions. Also NM KSAM files do not require the additional “KSAMK” type file to hold the indexes, all of the index pointers are now part of the NM KSAM file.

[Note: I’ll leave the final word on the subject to Stan Sieler: “Bug.”]