Hidden Value details
commands and procedures in MPE that can improve your productivity
with HP 3000 systems. Get a free NewsWire HP 3000 cap submit
your tip directly to us here at the NewsWire. To get your hat, send
your tips to email@example.com, or fax them to 512.331.3807.
Edited by John Burke
How is the maximum number
of jobs or sessions set by the LIMIT command related to the user
Duane Percox replies:
The maximum value you can
use in the LIMIT command for number of jobs and number of sessions is
set in SYSGENs MISC. It has no direct relationship to the user
license. In fact, not only can you set, for example, the limit on
sessions to a value greater than the user license limit, under some
circumstances you need to do just that. If you use RPM to create
sessions, these will show up in SHOWJOB under sessions and be subject
to the value set by the LIMIT command but do not count against the
user license. Think of the job and session limits set by the LIMIT
command as examples of resource management.
How can I give some Super
Users access to spooler commands without giving them OP or SM
Tom Brandt replies:
User.Account;commands=<command> allows User.Account to use the
specified command, but only if User.Account is logged on when the
ALLOW is issued. Once the user logs off, the permission evaporates.
ALLOW @.@ grants permission to all users whether they are logged on
I noticed there is a
parallel port on one of my cards in my HP e3000. How do I configure
it to work?
Steve Dirickson replies:
You cant. If you
try to configure it into SYSGEN youll get an LLIO error (6, I
vaguely recall) during startup. Which doesnt cause any
problems; its just an indication that the system cant
find anything to talk to the configured device. For no obvious
reason, HP e3000 no speak parallel.
I will be doing a system
INSTALL replacing all disk drives. The system has three user volumes.
The system will be backed up using the following store command:
STORE /SYS/, / -
If I restore using RESTORE
*T;@.@.@-log####.pub.sys;directory; olddate;keep;show=offline, will
all of the user volume directory information be restored as well? Is
it necessary to do a restore of only directory first?
Gilles Schipper replies:
Your command should work,
but I would restore the directory first and separately, as in
:restore *t;;directory;show.Then, check it out using :REPORT
I would also add
;create;partialdb to your restore command and use the
Posix syntax to ensure all your posix structures get restored. So,
the second restore should look as follows:
RESTORE *T;/ -
Now, all of this assumes
that you will have first re-created your user volume environment
using VOLUTIL and the appropriate NEWSET and NEWVOL commands
[Editors note: this
has been answered several times before, but it keeps getting
Please let me know how I
can reset my spoolfile numbers. Right now I am on o1066970. Can I
start it from 1 without rebooting the system?
Ron Horner replies:
The MPE setcounter
command will do the job. Just set your limit to the desired value and
the system does the rest. In your case enter the following:
counter=outsp;max=<desired max spool number>;base=1
If I have used the newlink
command and at a later date I want to verify that I did the command,
how can I see it? I tried showlink and that does not seem to
Donna Garverick replies:
fileset;seleq=[object=symlink] to list symbolic links.
Harry Morris adds:
If you want to see the
target of the link, use listfile linkname,7
Editors note: The
figure below shows an example:
How do I reset the HP
3000s internal support modem without rebooting?
Gilles Schipper replies:
You can do it with
SYSDIAG, but the module within sysdiag, CONSOLAN, requires a
suplicense code to run. Assuming you have it, the command required
consolan pdev=56/56 section=16
Your pdev number varies
according to your model of HP3000. 56/56 would apply to 9x8 models. I
believe the appropriate 9x9 pdev is 10/4/0.
[Editors note: if
you have a hardware support agreement, you can get the suplicense
code online from HPs support group at http://us-support.external.hp.com/]
We have a database that
has a field that is stored as a J2. In the Query manual J2 is
described as an integer value but it doesnt explain how it is
stored in the database. Is a J2 item packed, ASCII, or something
Wirt Atmar replies:
A J2 data item is
identical in format to an I2 integer, the only difference is that the
range is limited to an even decimal series (the largest
power of 10 that will fit under 2**31). However, IMAGE doesnt
enforce that rule. In fact, IMAGE enforces virtually no rules at all
about how data is stored in the database. Its essentially up to
the application to enforce any language (COBOL for example)-specified
artificial range restriction.
Ive got a database
message thats puzzling me. What does this mean and how can I
TURBOIMAGE RESULT AT
$0168a654; RETURN STATUS = 42
DBUPDATE, MODE 1, ON
LINE-CONTENT OF ??????????????
DBUPDATE WILL NOT ALTER A
READ-ONLY DATA ITEM
Dennis Heidner replies:
The password used to open
up the database does not give you write access to all fields. Your
program has used DBGET to retrieve a record, then modify an item
(which the password says you can view), and then DBUPDATE was called
to update the entry (perhaps with @; as the item list).
Correct the problem by:
1. Using different password OR
2. Fix program logic not
to modify the read only field in the programs copy of the data OR
3. Change program to use a
list of only the fields that truly need to be updated.
We use mirrored disks for
our non-system data. Last night I had the first hiccup since they
were installed. One of the disks went into a disabled
state. I put in a hardware call and began reading the Users
Guide. It instructed me to do a REPLACEMIRRVOL and call a CE if the
problem persisted. It worked. At least for now. What would cause a
mirrored disk to flip to a disabled state?
Chris Bartram replies:
Its been my
experience that Mirror/iX is very aggressive in flagging discs
bad (and un-mirroring them). Since all our drives are
under contract (and we keep a hot spare on site) when it drops one,
we pull it and swap it, then have HP come out and replace our spare
(dont like to take chances with them). [Note: All our mirrored
drives are in hot-swappable Jamaica enclosures.]
The few times we have just
enabled the drives back, we usually found that it would get dropped
We did have one continuing
episode of drives being dropped off frequently that turned out to be
because the original engineer that installed several of our discs had
put several instances of both discs (of a pair) on the same
controller. Under heavy loads the controller got overloaded and would
drop one of the mirrored discs. Also, beware running sysinfo, as it
is known to interfere with mirroring and can cause drives to drop
Bob Johnson adds:
Timing problems arise on
busy systems with Mirror/iX if interrupts are allowed to delay writes
between the mirrored drives. Placing the pairs on adjacent interfaces
greatly reduces dropping good discs.
Is there a way to get the MAC
(hardware) address of the NIC in a PC running an emulator connected
(NS/VT or telnet) to an HP e3000 from the HP e3000 side?
[Editors note: many
people replied, but Im going to use my editors
prerogative and use my answer.]
If the PC is on the same
subnet as the HP e3000, then the program nettool.net.sys
will show you the MAC address. That is, assuming it has had any
communication with the 3000 (a ping is sufficient to put the value in
the table). nettool.net.sys name;map;map;quit will list a
table of IP addresses and MAC addresses. If the PC is on the same
subnet, then the MAC address is that of the PC. If the PC is not on
the same subnet, then the MAC address is that of the router.
I do not know of any way
to find the MAC address of a PC on another subnet.
Where can I find a free
utility that would build scripts to restore my IMAGE/SQL
configuration after a detach?
Denys Beauchemin replies:
The utility name is
called DBTune/SQL (DBTSQL) and you can download it free from our Web
page at or directly at www.hicomp.com/dbtune.htm.
CSYs Jazz Web server also has a link to the above.
We have a system failure
each day when a cleanup job hits a certain spoolfile.
fscheck dutifully indicates corruption in the file label, and the
;FIX option seems to have no effect. And, yes, purging
the file via fscheck also crashes the system. How do I fix this?
Goetz Neumann replies:
If you are on Response
Center support, a knowing engineer can dial-in and mark the file
labels as unused/free and then run FSCHECK to delete the relating
directory entry. Another reboot and all should be okay. The
disadvantage of this method is that you will lose the
disc space that the extents of this file used. How bad that is
compared to rebuilding a volume set depends on your own judgment.