SolutionSoft Sponsor Message

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

We have several printers connected to a DTC and after rebooting the 947 which uses these printers, three LDEVs seem to be lost. If we do a SHOWDEV 127 this message comes up:
Logical device number 127 does not exist on the system.
How do I correct this problem?

Robert Schlosser replies:
If the printers were connected to a DTC, they are defined in NMMGR. Check to see if the DTC that they were attached to was not inadvertently removed from your configuration. If it was, and there were no other major changes made using NMMGR, you should be able to restore NMCONFIG.PUB.SYS to recover the configuration. A suggestion is to place the file in another group using the GROUP= option in RESTORE to play things safe. Check this file using NMMGR to be sure it is what you really want and then you can place it back in PUB.SYS. Go back into NMMGR and VALIDATE all of the subsystems used (NETXPORT and DTS/LINK), then restart your system with a START NORECOVERY.

Lars Appel adds:
As of MPE/iX 5.5, it might be sufficient to restart the DTS subsystem only:
:dtccntrl func=shutdown
:dtccntrl func=restart
You should only try this on a “quiet” system. Also, watch out for the pitfalls of online configuration vs. START RECOVERY.

[Editor’s note: The original questioner does not say what type of reboot was done. If NMMGR changes were done and implemented online, affecting the three LDEVs in question and a START RECOVERY was done before any START NORECOVERY, then none of the changes will be effected. MPE/iX uses nmconfig from the last START NORECOVERY.]

Is there a way to convert text to uppercase in a command file?

Glenn Cole and Tim Ericson reply:

ups()

Try :HELP FUNCTIONS for information on what functions are available. Help is also available for each individual function (e.g., :HELP UPS).

Our 969 had a hung session that we could not get to go away. This morning we tried shutting down the system to kill the session. The shutdown process has stopped at “Shut 3” and will not go further. What is the next step to try?

John Burke replies:
CTRL-B followed by RS. I’ve seen the same thing on our system. Sometimes these hung sessions will just not go away —even during the shutdown process.

I’ve got what I thought was a simple backup job stream. The first one JBACKUP runs and generates all the full backups for the week that will run each night. It also re-runs itself (in theory) every Monday at 10 A.M. Come to find out after coming to work today that it streamed approximately 84 copies of JRRALL and one JBACKUP. Why is this happening?

David Rutherford, Robert Schlosser and David Birnet all reply:
The problem is your STREAM JBACKUP.OPERATOR.SYS; at=10:00;DAY=MONDAY command. Because this job is actually executing this command at 10 A.M. on a Monday, it will physically stream this job NOW, not next Monday. The streamed job will also stream this job again, as many times as possible until 10:01 A.M. is reached, when it will correctly stream it for next Monday at 10 A.M. A solution would be to put a PAUSE 60 before the streaming of the JBACKUP job for next Monday.

What default login accounts are initially installed on an HP 3000? What are the default permissions?

[Editor’s note: I know of OPERATOR.SYS and MANAGER.SYS.]

Lars Appel replies:
See the SUPACCT.PUB.SYS job that is streamed during install or update.

We’re neophyte HP 3000 users, and while migrating our 10Base2 network to a 10BaseT network, we noticed that the HP 3000 and DTC are generating a large number of packets (between each other?), albeit very small ones. Overall utilization on each port is only one percent and the traffic is approximately 140kbps and 30 packets per second. This occurs even when there is nothing on the network. Why?

Fred Metcalf replies:
Make sure all your transceivers have SQE set “on.”

The original questioner has added a summary of advice:
SQE is also known as the ‘heartbeat’ and is intended to tell the network card that the network is still alive. Some equipment (including the HP 3000 and DTC apparently) prefer this ‘heartbeat’ turned on. In the case of the HP 3000, if the DTC doesn’t have SQE turned on, it sends approximately 30 alerts/second to the HP3000 saying it has a problem and the HP 3000 (usually, unless configured otherwise) logs these alerts to disk, leading to a slow down in performance of the HP 3000. If you suspect this may be a problem on your HP/DTC, use the command:

:linkcontrol @;status=all

The last line gives information on the number of Heartbeat losses, which should normally be zero. Ours was in the hundreds of thousands.

We’ve connected a 2564B printer (with a dual interface card) via a Centronics parallel interface to a JetDirect external box. However, we cannot print a test page from the JetDirect. The JetDirect works perfectly well with a LaserJet and the same cable. The interface card has also been swapped out, but the 2563B is still dead.

Jim Brust replies:
The test page can be generated in not only PCL format, but also ASCII, PostScript, or HP-GL/2 depending on how the unit is configured. How to configure was at www.hp.com/cposupport/networking /support_doc/bpj02590.html

“The LEDs and test button are used to select a PDL for the configuration page. When holding down the test button after three seconds the LEDs begin to cycle through four configurations for the PDLs …”

PDL	Status LED	Activity LED
HP PCL	Off	Off
ASCII	Off	On
PostScript< /td>	On	Off
HP-GL2	On	On

[Editor’s note: Here is another tidbit on JetDirect external boxes: to reset to default values, unplug it, hold down the TEST button and plug it back in to a power source.]

We have maxed out our VTSERVER sessions and I am drawing a blank on where to increase this setting. What’s the command?

Lee Gunter, Robert Schlosser and Jeff Woods all reply in short order:

:NSCONTROL SERVER=VTSERVER,,<newmaxservercount>
[Editor’s note: this takes effect dynamically.]

Kevin Miller added:
You can edit the file VTSERV.NET.SYS. In my case lines 7-8 have a 0 and 300. I changed them to 20 and 400, respectively, streamed JCONFJOB.NET.SYS then
NSCONTROL STOP/ABORT
and finally,
NSCONTROL START
My default VTSERVER is now 400 max, 20 reserved.

I’ve been asked to rename a DTC currently managed by Openview DTC Manager. The directions provided are quite clear, but I’m wondering is there a relationship, based on the name, between DTC Manager and NMMGR?

John Skelton replies:
It depends which name you are referring to. The eight character DTC name is just a label; however, the three part (node.domain.org) DTC NODE name needs to be the same in NMMGR as it is on the DTC Manager workstation.

If I use the Posix commands to move or copy a file, the moved file ends up with an ACD on it when it did not have one in the original account. Is there a way to eliminate this behavior or to remove the ACD under Posix?

HP’s Jeff Vance replies:
This should not happen with a CI COPY command, but you do get an ACD using the shell’s cp command. However, I am able to delete the ACD after doing a shell cp.