Click here for The Support Group Sponsor Message

Hidden Value (June 1997)

Compiled by John Burke

How do I configure a tape drive for AUTOREPLY?
Michael Berkowitz replies:

Use the :SYSGEN command and then do the following:

sysgen> io

io>md ldev=xxx mode=AUTOREPLY

io>ho

io>e

sysgen>keep

[Editor's note: This procedure will work on MPE/iX 4.0 through 5.5. If you have more than one tape drive in the same device class, you will need to specify the LDEV number in your file equation for AUTOREPLY to work.]

We just purchased two 4.3-Gb disk drives. After they are configured, should I do a reload? If I do a reload, how do I distribute the files evenly across all drives?

Stan Sieler replies:
No, a reload isn't needed. The only reasons I can see to do a reload in your case are:

1) You've decided to change the number of volumes in one or more existing volume set and are taking this opportunity to do it. For example, if you go from a 10 disk MPEXL_SYSTEM_VOLUME_SET setup to a 2-disk MPEXL_SYSTEM_VOLUME_SET, 5 disk USER1, and 5-disk USER2 volume sets.)

2) You want to "balance" the space available across all of the disks in the volume set(s) that the new drives are added into. Two products are available that can re-balance an existing system. [Editor's note: De-Frag/X from Lund Performance Solutions and Disk Space Manager, from Bradmark Technologies]

If you reload the files yourself, you're likely to run into future FSCHECK problems because some files that are expected to be on LDEV 1, but not required to be there, may have been restored onto an LDEV other than 1. This problem generally shows up with FSCHECK finding entries in the label table for files in the DIAG.SYS group without corresponding entries in the directory. I'd recommend at least restoring @.DIAG.SYS onto LDEV 1. [Editor's note: SYSSTART is an example of a file that must be on LDEV 1.]

Is there some way for a job to do a recall and automatically reply to the request? I have 11 replies every night for the system backups, and I would love to automate this so that I do not have to continuously do replies all night long.

Assuming you don't want to configure the drives for AUTOREPLY, Carl W. Kemp replies:
Do yourself a favor. Go get a copy of MasterOp from Allegro's Web Page. It is FREE and has all sorts of reply handling capabilities, as well as being a pretty powerful scheduler (okay, so being the author, I am a little prejudiced). Then the command from MasterOp would be:

WHENEVER (REPLYINFO(#J@,TEXT)="*LDEV*") DO &
:REPLY ~[REPLYINFO(#J@,PIN)],ldev

Which release will have an item for the file creation time in FINFO? I thought it was going to be 5.5.

Siegfried Gladitsch of HP replies:

It is not part of 5.5 directly. It was planned for patch MPEHXY8, which contains the LISTF,ACCESS enhancements. However, it did not make it. The functionality should definitely be in MPE/iX 6.0.

For the third time in three weeks HP Predictive Support has informed me that logging has been disabled and the system needs to be rebooted to re-enable it. Has anyone else encountered this? We are on MPE/iX 5.5.

Lee Gunter replies:
Yes, we've had it occur on both our systems since 5.0 (we're now on 5.5, with no fix yet). The problem is documented in SR# 5003312272: LOGFILE opened with MAPPED ACCESS and was originally identified by HP labs as a problem with Vesoft's %VEAUDIT LISTLOG command. However, Vesoft now says that HP has acknowledged the problem as theirs. Vesoft can also reproduce the problem with their %BACKG STOPJOB/STARTJOB commands (SR# 4701326348: FCLOSE of COPY-mode message file). Any of these tools could cause your problem. HP provided us with several DEBUG command scripts to

1) Set a mask that can prevent logging from disabling (XEQ at system startup time),

2) Re-enable logging without rebooting the system, and

3) Turn the log-disable prevention mask off. I suggest that you contact the Response Center and reference these two SR's and see if they'll provide you with the scripts.

I just purchased 3 DDS-2 dat drives for my MPE 5.0 HP3000 system. I want one of the drives to be configured with compression off. How do I control the compression setting for DDS-2 drives?

Per Ostberg replies:
Check out the script devctrl.mpexl.telesup that interfaces to the program devtool.mpexl.telesup. Syntax for controlling compression is:

:devctrl.mpexl.telesup ldev# compression= [ENABLE|DISABLE|NOCHANGE]

[Editor's note: devctrl can also be used to control EJECT behavior and, as of 5.5, LOAD behavior (See the MPE/iX 5.5 Communicator for details).]

We currently use TurboStore in parallel with hardware compression on four DDS-2 drives to backup 34Gb of disk from a 959/400. It takes 2 hours. What kind of time can I expect to save with the new DDS-3 drives?

Gary Dietz replies:
I don't use DDS-3 (yet), but according to the literature I have from HP the specs for DDS-3 are 12 Gb native capacity, 24 Gb at a maximum 2-to-1 compression, and a 2 Mb/sec transfer rate (at 2-to-1 compression)

Since the transfer rate of the 4/8-Gb DDS-2 drives is 1 Mb/sec (at 2-to-1 compression), the DDS-3 should theoretically be twice as fast as the DDS-2 (if your system can keep the data flowing to the tapes fast enough). Or you could always look at DLT-40, which gives 20Gb native, 40Gb compressed capacities with a 3 Mb/sec transfer rate.

I was hoping to install a couple of patches this weekend but am having trouble getting them onto my system. I downloaded the patches from http://us-support.external.hp.com and also downloaded UNPACKP. When I run UNPACKP, carefully following the instructions, I get the following error message:

Creating directory(ies) in path: 'L00JXJ4A'
1461 128W 396 396 -rw-rw---- L00JXJ4A
Unrecognized token '#223'.
END OF PROGRAM
Unknown failure. Quitting script.
DUPLICATE PERMANENT FILE NAME (FSERR 100)
rename UUOJXJ4A,MPEJXJ4A
RENAME failed due to file system error. Not renamed. (CIERR 373)

What does it all mean? What am I doing wrong? Am I going to get Sunday off after all?

You are probably using an incompatible version of MOVER. Unfortunately different versions of MOVER having the same version ID A.00.02 are floating around. You need to use the version of MOVER with eof=635. It is available at http://jazz.external.hp.com.

[Editor's note to HP: This question comes up nearly every month and is an obvious source of frustration to many users. Please include a note in the patch section of us-support.hp.com about this. Better yet, include a downloadable version of the correct MOVER, or at least a link to the version on Jazz.]

How can I look at the contents of a message file without emptying it?

Use FCOPY and the COPY keyword for FILE as in

FILE M=msgfile;COPY

FCOPY FROM=*M;TO=...

Lee Gunter and Gilles Shipper reply:
If you use GL capability, you don't need group passwords, and that's what's so great about GL. By assigning each group security as follows, (R,L,X,W,A,S:GL),and giving all users GL capability, you have the security you need without the nuisance associated with group password maintenance. (You were going to regularly change the group passwords, weren't you?)

This is due to the very useful aspect of GL -- one can only be GL with respect to one's own home group. Once you stray from your home group, you lose your GL capability. Of course, the AM user has access to all files in the account.

How many hours does it typically take to update from MPE 5.0 to 5.5?

If all goes well and you are updating from a CD-ROM, the update should take two to three hours -- not including the pre-upgrade time (validating tapes, etc.) if you have no reactive patches. The update takes closer to 4 hours if you're updating from tape and about six hours if also applying PowerPatch and/or additional patches.

We recently put an HP 3000 server on our network for telnet access and network printing. We have a lot of Epson 1070 printers that we want to hang off of HP JetDirect EX+ printer servers. Why is it that if I hang an HP printer (ie. LaserJet, DeskJet) on the HP JetDirect, everything works fine, but I cannot seem to get the Epsons to work correctly?

HP's Larry Byler replies:
You have ignored one of the following ground rules for network printing using the MPE-provided network printing package:

-- TCP/IP network connection (probably okay, although you don't say)

-- JetDirect interface (check)

-- PCL-speaking printer (Bzzzt)

The spooler itself uses PCL sequences to manage its printers. Bottom line: your Epson printers will not work with the MPE-provided network spooling package unless they speak PCL. You will either have to configure them on a serial link, buy PCL printers, or avail yourself of a third-party solution that does not require PCL printers. [Editor's note: This is an excellent place to turn to PRINTPATH/SE from RAC Consulting (360.357.9572 or), reviewed in our TestDrive of May, 1997, or Minisoft's (800.682.0200) NP92.]

How do I get rid of "hung" VT sessions?

HP's James Hofmeister replies:
1. nscontrol killsess = #s.... This command is not documented in the 'help nscontrol' or in HP manuals. This tool sends a TCP RESET to the VTSERVER TCP connection.

2. showconn & abortcon: These unsupported utilities are located in NSRV0000.TELESUP and delivered with 5.0 General Release NSS patch. This tool sends a TCP RESET to any TCP connection specified by a process pin number.

3. Connection Assurance Timeout default timeout is 40 minutes in the case of a client which powers down.

How can I delete groups of spoolfiles matching certain criteria?

Michael Berkowitz replies:
Try this command:SPOOLF O@;SELEQ=[your criteria];DELETE;SHOW

We are installing an HP 3000 in another building. It needs to be connected to the existing 10Base2 network in this building. We are planning on running fiber-optic between the buildings. How do I make the transition on each end of the fiber to the 10Base2, a hub or a converter?

Chris Hill replies:
Most hubs allow for a fiber transceiver to be inserted. HP's hubs require you to remove the AUI card and replace with the transceiver. This allows the fiber to be connected to the hub, which in turn is connected to the 10base2 network.

Is there a way within MPE to purge non-database privileged files (files with a negative file code) located in HFS space?

Use these commands
:FILE X=$NULL
:STORE privfile;*X;SHOW;PURGE

(Another approach suggested by more than one individual was to use the utility FSCHECK which is part of the TELESUP account.)

We are trying to slave print to a PC with Reflection 5.2 using an NS/VT connection to the 3000. The only way to flush the buffer seems to be to hit the 'CLOSE PRINTER' button. How can I configure Reflection to do this automatically? We cannot do anything programmatically since we are using a third party package that writes reports to the printer. This was not a problem with terminals or the DOS version of Reflection.

Mike Gabriel replies:
1. Start Reflection

2. Under File, choose Print Setup

a) Mark the box labeled Bypass Windows Print Drive b) Mark the box labeled Disable Printer Translation c) Click on OK

3. Double-click on the bottom bar of the Reflection window until it changes to RB Command and type .PRINTERTIMEOUT = 10

The value of 10 for PRINTERTIMEOUT is in seconds. After this amount of idle time, Reflection will close the printer and print the output.

We seem to be getting a "meaningful" descriptive message on the console when a tape job is finished. The jobs run fine, and the tapes validate just fine. The drives are a mixture of both DDS-2 & DDS-3. Here's the message; can you supply the translation?

MISSING MSG. SET=1. MESSAGE=417.
PARM(S)= 8;;;;
GENMSG CALLER STATUS: %101000 DELTA-P: %177777 GENMSG CALLER+1 STATUS: %140023 DELTA-P: %040000

Mark Bixby replies:
I opened an HP Response Center call about this last December right after I upgraded to 5.5. It's a known bug caused by a few missing messages from the message catalog relating to tape quality and/or the need to clean the drive. [Editor's note: According to the Response Center, the missing messages -- 415, 416 and 417 -- mean replace the media and/or clean the drive.]

Copyright 1997, The 3000 NewsWire. All rights reserved.