Q: How can I determine
the validity of an SLT tape?
A: Bob J and Andreas
Schmidt replied:
Use
CHECKSLT.MPEXL.TELESUP option 1.
Q: Is there a way to
view deleted records in a Native Mode KSAM file? "NOKSAM" which
worked for CM files does not work.
A: Ken Hirsch replied:
If you open the file
with the COPY aoption, then you can read all the records. There is a four-byte
field at the start of each record that indicates whether or not it is deleted.
This almost works:
file
ksamcopy=yourksam,old;copy
fcopy from=*ksamcopy;to=
If you try this, you'll
see that FCOPY reads by the original record length and gets off by four bytes
each record. If you write a program and
read by the larger (by 4) record length, it will work.
Q: Does anyone have an
example of using TAR to put data on to tape, on a
HP3000?
A: Rick Clarke replied:
This is from the HPITRC
Document ID KBRC00000361
1)Create a tape node
:MKNOD "/dev/tape c
0 7"
2)Enter posix shell
:SH -L
3)Mount blank tape and
enter the tar command
shell/ix>tar -cvf
/dev/tape /ACCOUNT/GROUP/FILENAME
Lars Appel added:
If you do not want to
use it from the Shell, but from the CI...
:xeq /bin/tar "-cvf /dev/tape
/ACCOUNT/GROUP/FILENAME"
Note, however, that only
the Shell will expand wildcards...
:xeq /bin/sh " -c 'tar -cvf /dev/tape
/ACCOUNT/GROUP/FILE*' "
Oh, and the :xeq above
is optional as well.
Q: We recently put in
place a VPN in order to reach our HP3000 from outside our firewall (E.g., for
telecommuting). We have had good success with the connection except when
attempting to use REFLECTION file transfer. The transfer will timeout, but, and
here's the interesting part, not all of the time. Every once in a while, after several retries, the transfer
succeeds. Has anyone else experienced this problem?
A: Bob Sorenson replied:
Try setting "Fast
file transfer" to "Never".
This will be found under the WRQ tab. It should solve your problem.
Editor: Normally I do
not include Q&As that deal with 3rd party products. However, because this
speaks to an important feature of MPE/iX, I am including it.
Q: If I say 'altfile
<somefileset>;xltrim' in mpex, what really happens? Is mpex working
directly on each file or does it make a 'work' copy?
A: Ron Horner, Paul
Christidis and Erik Vistica replied:
MPEX merely (ignoring
for a moment all the verifications that it does) opens up the file and closes
it with the option that returns to the system any 'unused' space on the 'last'
extent of the file (NO second file). So a file with 10 equal extents, for
example, will end up with 9 equal extents and 1 smaller (if there was any
'empty' space). This became 'achievable' as a result of how disc space is
allocated in MPE/iX.
One thing that you
should keep in mind is that the 'mod date' of all 'xltrimmed' files will be
changed (unless ';keepamdates' is used) and that may impact your next backup.
It bit me the first time that I used it because my backup needed another tape
and thus I got paged in the middle of the night, etc.
Q: Today I had a perfect
application for the NEWJOBQ feature in MPE/iX 6.5. We have two groups of users.
One group submits jobs that take about 30 seconds each. Typical jobs for the
other group take about 5 minutes each. So I thought I’d give the second group
of users their own job queue.
NEWJOBQ ALTJOBQ;LIMIT=1
LIMIT 1 (for HPSYSJOBQ)
Now I submit a long job
into the ALTJOBQ queue, and a quick job into the default job queue. The second
job goes into the WAIT state. Why?
A: Michael Berkowitz,
David Darnell and Denis St-Amand replied:
Your NEWJOBQ statement
is correct, but your second statement didn't do what you thought. To put a
limit of one on the HPSYSJQ job queue, your statement should read :LIMIT
1;JOBQ=HPSYSJQ. By saying :LIMIT 1, you
are changing the total job limit on the system to one. Since the total limit is
one, and the long job in ALTJOBQ is still running, the second job waits even
though he is the only job in his queue.
Q: What is the command
to abort a hung session? I tried
ABORTJOB #s3456
I seem to remember there
is a command that will do more?
A: John Burke replied:
=SHUTDOWN
But seriously, there is
a chance that if it is a network connection,
NSCONTROL
KILLSESS=#S3456
will work. If it is a
serial DTC connection, ABORTIO on the LDEV should work. Finally, depending upon
what level of the OS you are on, look into the ABORTPROC command. This might
help as a last resort.
Otherwise, I'm afraid
you are probably stuck with having to do a reboot. Hung sessions happen much
less frequently these days but are still one of the worst things short of a SYSABORT
that can occur to a production system.
[Editor's note:
NSCONTROL KILLSESS worked for the questioner.]
Q: What does
HPSWINFO.PUB.SYS show? All software or
only installed software? So, how do I know what HP software is installed?
A: John Burke replied:
Generally speaking,
HPSWINFO.PUB.SYS is a record of system software level and patching activity.
If you want information
on HP software installed then you want to run psswinvp.
On 6.0, the calling
sequence is
run psswinvp.pred.sys;xl='xl.pred.sys,xl.diag.sys'
On 6.5, the calling
sequence is
run
psswinvp.pred.sys;xl='xl.pred.sys,diagxl.pred.sys'
On 7.x? Sorry, someone
else will have to chime in.
There is a command file
(psirpt.pred.sys) but is is broken on MPE/iX 6.5; hence I've given the explicit
calling sequence.
Q: I am in San
Francisco, CA. Should my TZ variable be PST7PDT or PST8PDT?
A: Steve Dirickson and
David Powell replied:
Not sure where else it
is documented, but there is a list in the 1991 edition Cobol manual. PST8PDT is
it.
Q: Can anyone tell me
what this image error is and how to fix it?
TURBOIMAGE RESULT AT $0008b4f8; RETURN STATUS
= 44
DBDELETE, MODE 1, ON #34 OF HEALTH.DATA
CAN'T DELETE A MASTER ENTRY WITH NON-EMPTY
DETAIL CHAINS
A: Jonathan Backus and
Ray Shahan replied:
You are trying to delete
a record from a dataset that appears to be a Manual Master. The problem is
there are records in one or more associated Detail datasets that have this same
key value. Therefore, there are "non-empty" chains and IMAGE refuses
to delete the master record. You have to delete all of the chained records from
the associated detail datasets first.
Q: I'd like to synch the
time on my 3000 with my NT40 network. The NT side does regular, automated synch
to NIST.
A: Mark Bixby and Jeff
Kell replied:
First, insure your
timezone is absolutely correct (:setclock/:showclock) and you have a system
logon UDC to setvar TZ to the correct timezone.
Install NTP (available
on Jazz if you didn't get it otherwise) and use the 'ntpdate' function to sync
your clock to the NT servers. Suggest
you do this in a batch job that issues the ntpdate command and then :STREAMs
itself;IN=xx to periodically perform the synchronization.
Q: I want to set a
numeric variable, put a alpha in front of it, run a program and increment the
numeric variable by 1; i.e.,
setvar num 123
setvar alpha 'r'
setvar composite
!alpha!num
....
run program
setvar num num+1
setvar composite
!alpha!num
Setting the composite
variable fails.
A: David Burney and
Friedrich Harasleban replied:
Instead of :
setvar composite
!alpha!num
try
setvar composite
"!alpha!num"
Q: The LISTDIR option of
RESTORE applies only to NMSTORE tapes. It cannot be used for MPE/V format tapes.
What utility can I use to view the file list of a MPE/V tape?
A: Gilles Schipper
replied:
VALIDATE, from the
contributed libarary. There should be a copy in the TELESUP account.
Q: If I want to
reconfigure several disk drives so that I'm using three, instead of four, FWD
SCSI cards, do I need to perform a system reload? Only the drive paths would be
changing. All affected drives are on the System Volume Set and I would not have
more than eight drives on a path.
A: Guy Paul replied:
Just did this last
weekend. If all you are changing is the path, then no, you do not have to
reload any data.
You can even change the
ldev number without reloading, EXCEPT as I recently discovered the hard way, if
the ldev you are changing is part of mirror.
Q: While viewing logs in
sqlutil it does not seem clear how to view how full the log file is We have had
problems whereas the logs have reached capacity and no new logs were available
to change to. Any suggestions on how I can see the true size of the file:
A: Michael Gueterman
replied:
You can use SQLUTIL's
SHOWLOG command to find out how much "usable" log space you have
left. Near the beginning of the output, there is a line that reads:
Number of Free Block(s):
xxxxxxxx
That xxxxxxxx value will
be the total amount of free log space across ALL defined log files. If you are
using ARCHIVE logging, then you must store that log file before the space
becomes available again. If you’re using NONARCHIVE logging, then, as your
transactions are committed, the space in the log file becomes available. If
you’re getting a LOG FULL condition, then either you need more/larger log
files, and/or store the log files more frequently.
There is a chapter
called 'Calculating Storage for Logging' in the Allbase/SQL Database
Administration Guide that goes into detail about how to determine appropriate
sizes.
Q: Is it possible to
list the files that were stored on a specific CSLT? If so, how?
A: John Burke, Gary
Jackson and Paul Courry replied:
CHECKSLT.MPEXL.TELESUP
(option 1) will list the system files.
VSTORE can be used to
list any "store format" files.
Q: Is there any way to
list previous user log-ons? We need to know who was logged onto the system last
night.
A: Andreas Schmidt,
Jonathan Backus and John Clogg replied:
Use LOGTOOL. If you are
logging event type 102 (Job/Session initiation) or type 115 (Console messages)
you can find what you need quite simply.
If you aren't familiar with LOGTOOL, just run it and type HELP LIST.
There are also CSL tools that will help format the system logs.
Q: I've been asked how
the HP3000 resolves IP addresses to DNS names by someone in Networking. His
query is: "Can it be pointed to a DNS server or does the name resolution
have to be hard-coded in a HOSTS file?"
A: Guy Paul replied:
Actually both are now
possible with the latest NSR patch available for 6.0 and greater releases. This
will allow the NSSWITCH functionality that UX has had for years and allow you
to use the reslvcnf.net and hosts.net solution and specify which one to try
first.