Q: I've been experimenting with setting up
the FTP service inside JINETD. I've got it up and running OK, but I'm having
problems applying the security I think necessary for our site.
I've had problems when trying to configure
INETDSEC.NET.SYS. I want to ensure that the FTP server on each HP3K will only
accept connections from the other HP3Ks (and itself) in our site, but a problem
occurs when testing FTP in loopback mode.
I have added the following line to
INETDSEC.NET.SYS on both HPs
Allow 123.456.789.0 987.654.321.0
With this line in place each HP can connect
to the other HP via FTP, but is refused a connection (Ftp error 9) when I try a
loop back test. If I remove the "allow" line in the INETDSEC file,
then the loopback test is successful; i.e. I can successfully connect from any
HP to any HP. But it also opens up the HP to any remote client, which isn't
what I want! Any suggestions?
A: Gerhard Roets and Gary Paveza replied:
Add 127.0.0.1 to the "allow" line
as well.
Q: I understand when using the SETCOUNTER
command with a MAX value that once MAX is reached, it starts back at BASE. For
example, doing a SETCOUNTER OUTSP;BASE=1;MAX=9999 would start the output spool
file numbers over at 1 and set a maximum value of 9999. At 10,000, it starts
back at 1 or at the first available number after 1. But what happens if all the
numbers between BASE and MAX are in use?
A: Jeff Vance replied:
I have access to code but no time to look
right now. My guess is that the counter will begin at MAX+1.
However, we also have to handle the case
where all available numbers are taken, up through the absolute maximum. In that
case you'd probably get a JSMAIN error related to being out of system
resources.
John Burke added:
OK, I went ahead and tested it for
everyone's edification. Jeff's guess is correct. If the spooler runs out of
available numbers below MAX, it just keeps going with the next available number
above MAX.
Q: I have jobs that are in a WAIT state. The
JOBFENCE is at 2 and the INPRI of the jobs are 8. What can I do?
A: James Reynolds replied:
Issue the OPENQ @ command.
Q: Does anyone know the recommended life or
max usage of DLT tapes?
A: Denys Beauchemin replied:
www.dlttape.com will have lots of
information for you. One thing to remember is that when they talk about the
number of passes (500,000 passes), it does not mean number of tape mounts.
Talking about SuperDLT tapes, the tape is
divided into 448 physical tracks of 8 channels each giving 56 logical
tracks. This means that when you write
a SuperDLT tape completely you will have just completed 56 passes. If you read the tape completely, you will
have done another 56 passes.
The DLTIV tapes (DLT7000/8000) have a
smaller number of physical and logical tracks, but the principle is the same.
The number of passes for a DLTIIIXT and DLT
IV tapes is 1,000,000. It's 500,000 for
the DLTIII. The shelf life is 30 years
for the formers and 20 for the latter.
Q: I have a Master data set that is rather
large. In order to improve the rather high number of secondaries, I thought I
might try turning on MDX. But I seem to remember that this can't be done on
Jumbo datasets. What is the story?
A: Rene Woc replied:
MDX on master datasets larger than 4
gigabytes is available with Large File Datasets (LFDS). LFDS is part of the
TurboIMAGE C.10.04 enhancements on MPE/iX 7.5. You cannot use MDX on jumbo
datasets.
Q: We have recently received following
messages on the console when booting our system:
MPS: installed device clone as major 1
MPS: installed device sad as major 121
MPS: installed device pipe as major 137
MPS: installed module sw2proc
...
Does anyone know what these MPS messages
stand for?
A: Andreas Schmidt replied, quoting HP
Support Pages:
These messages appear on the console the
first time pipes are used. They are generated by the system, not the POSIX
shell and utilities. The message is normal and only occurs once per system
boot.
Gavin Scott added some detail:
I would hazard a guess that "MPS"
is "Mentat Portable Streams", the SYSTEM V STREAMS library that HP
licensed for use in implementing named pipes and similar functionality in MPE.
These messages will always be produced the first time after boot-up that
something invokes the Streams subsystem by using a named pipe etc.
The MPS code in MPE is one of the few
components of the operating system not entirely owned by HP, and as such it is
one of the complicating factors for any future Open MPE initiatives or
alternative MPE licensing.
The MPS startup messages are normal and you
should ignore them.
Q: We have an N-220 1 CPU. The current card
runs several DTCs, which limits us to 10Mb. If we were to add a second card,
leave the DTCs and serial devices on the first card, and move the pure network
connections to a second card at 100mb, are there any down sides we should know
about? We are currently running MPE/iX 7.0.
A: Robert Schlosser and John Clogg replied:
You can configure the network card to 100mb
and place the DTC on a switch with its segment running at 10mb.
Q: We just added a gateway to our HP3000. I
believe there is a way of refreshing the system's networking without re-booting.
Does any one know the command?
A: Jeff Kell and
:NETCONTROL NET=netname;UPDATE=ALL
Specify your own configured 'netname',
usually LAN1.
Q: Evidently Query will only allow me 256 characters
as output using the REPORT command. Anyone know if I can change this limit?
A: John Hornberger replied:
I just checked on docs.hp.com and found out
that 256 is the design limit for output line length. To view all of Query's
other limits please check the following URL:
http://docs.hp.com/mpeix/onlinedocs/32650-90889/32650-90889.html
Expand the contents listing for
"QUERY/3000 Specifications" and then click on the "Limits Used
By Query" subitem.
Paul Scot suggested a workaround:
You can accomplish this by
"splitting" your output onto two or more lines, then deblocking the
output with FCOPY after it is written to disk.
Q: I'm having a problem with what I think
should be a very simple thing, reading and writing 8 byte real numbers. I have two fields in one data table defined
in IMAGE as R4 and defined in Dictionary as R(28,6,8). When I read existing
values in this field they don't reflect the same values as shown by Query. When
I insert a record with known values into the table I don't see the values I set
reflected in the data when view by Query.
A: Michael Anderson and Steve Cooper
replied:
There are two kinds of real numbers on
MPE/iX systems: HP 3000 reals (aka compatability reals) which existed on the
classic HP 3000 systems and are designated in IMAGE as R4, and IEEE reals,
which are designated as E4. Generally
speaking, CM programs play with R4s and NM programs play with E4s, although
there is an intrinsic that will let you switch back and forth between them,
possible losing some precision along the way.
The way you describe the problem, if I had
to take a wild stab, I would guess that your programs are really using IEEE
reals, but you lied to IMAGE and said they were HP 3000 reals. IMAGE couldn't
care less, but Query is now confused.
It can't tell the difference; bits are bits. If that's the case, change
the field type, but be careful not to do any conversion of the fields
themselves.
Q: We need to prevent update access to small
subset of datasets of a large database during a limited period without at the
same time preventing read access. The bulk of the database should behave normally.
A: Mark Wonsil replied:
You need to change the dataset security with
a third-party database maintenance tool.
Q: On Samba 2.0.10 I'm seeing a new oddity,
the Samba JSMAIN process goes busy for 30 seconds, then goes idle for 30
seconds, then goes busy, etc. Below is an example taken from the display of
Glance on an otherwise idle system. Any ideas?
HPB1787 B.11.10 HP GlancePlus/iX 13:05:41 Global
...
A: Guy Paul, Lars Appel and Denis St-Amand
replied:
Based on you version of Glance, B.11.10, you
need to upgrade your Glance version to B.11.14 via patch GLAMX80C for 7.0.
There are known issues with your version of Glance reporting bogus JSMAIN
garbage. This is a Store-only patch so can be installed without downtime.
Q: I recently acquired an HP3000 937LX, with
no video output, but only a mini-DIN port for attaching a serial console. Can
anyone direct me to the pinouts or an adaptor for this port?
A: Alfredo Rego replied:
I made a cable for my Mac PowerBook thus:
DB9 to MD6 (male)
3
3
2
4
5
1
If you use a PC, you can plug the DB9
connector directly to your serial port. On my Mac, I use a USB adapter. The MD6 connector goes into your hp3000's
console port.
Larry Liska added the information for a DB25
connector:
Pin out for standard console cable
(A1703-63003) is:
DIN CONNECTOR
25 PIN (MALE)
FRONT VIEW
___________________
2 1 3 ______________ 2
\ /
4 __ 3
\_______________/
6||5 4 ______________ 3
^
Belden Cable 9819 28 gauge
| 1 ______________ 7 specified in the HP manual
|
KEY
Q: I have a file that has nulls that need to
be stripped. Is there a way in Quad to do this?
A: Jim Kramer and
To strip nulls using Quad you would say:
/c %0 to :: all
or (shorthand)
/c %0 :: a
Using hex and octal in strings is covered
under the Strings topic in Help.
Q: We recently downloaded and installed
text2pdf on our HP 3000. The installation went fine and we can create pdf files
in the POSIX shell as long as we are signed on as MANAGER.SYS. Our intentions
were to use this under MPE (we are running MPE/iX 5.5) and creating some of our
batch reports in pdf format and emailing them. What do we need to do (is it
possible) to access text2pdf from other users so we can create pdf files in our
jobstreams?
A: John Burke and Mark Wonsil replied:
Simply copy it to some group with
appropriate capabilities. PUB.SYS works, but I would not recommend it. A better
choice would be to copy it to a group and account where you keep miscellaneous
utility programs. There are no specific capabilities required for text2pdf.
Q: I'm curious to know if the Store-to-Disc
option allows for a 'parallel store', analogous to a parallel store to tape.
For example, when a parallel store to tape is performed, the store writes to
two or more tape drives at the same time. Is there a parallel store-to-disc
option that allows for the store to write to two or more disc files at the same
time (as opposed to running multiple store-to-disc jobs)?
A: Gavin Scott and Joe Taylor replied:
Yes, the same syntax for parallel stores
works for disk files as well as tape files. I really don't know if you would
get any benefit from this, but if you went to the trouble of building your STD
files on specific disks, then it might be worthwhile.
Q: It seems that JINFO only works against
jobs that are either executing or waiting.
If I do "CALC JINFO('#J1234','EXISTS') it comes back TRUE for
executing and waiting jobs but if #J1234 is in the 'READY' state it comes back
false.
while jinfo('!hplastjob','StdlistSPstate')
<> "READY"
echo processing...
pause 5
endwhile
I'm trying to wait for the job to finish but
if the job is already finished by the time it gets to the loop it fails with an
error (Job does not exist. (CIERR 3042)). This, even though the 'JINFO'
documentation states that READY is one of the possible responses for this
option. Anybody have any thoughts?
A: Wyall Grunwald and Jeff Vance replied:
JINFO also works on scheduled or
:breakjob'd. If you don't see the job/session via :showjob then JINFO will not
see it either.
In your loop, there may be a window where
STDLIST output is READY and the job still exists (probably small if the
condition exists at all). More likely, the documentation should be changed to
remove that state, even though READY is a valid spoolfile state.
The PAUSE job=#Jnnn command is a good way to
wait for a job to terminate. If the job is already done, the PAUSE command
returns immediately. You can also place a maximum number of seconds to wait for
the job to complete and PAUSE will return once the job finishes or that max
seconds expires, which ever occurs first. I believe that PAUSE will be more
efficient than JINFO in a WHILE loop.