As an alternative to SLEEP, use PAUSE. A process in PAUSE will return
the call stack via PORT.STATUS.
SLEEP, CALLC, PCPERFORM, and waiting on a record lock are some of most
common conditions which will not return a call stack via PORT.STATUS.
Dean Armbruster
System Analyst
757-989-2839
-Or
Yes, we did change ours to -m3. Waking within three seconds of the
desired time is good enough for us. Ten seconds was too far off, and we
were concerned about cleanupd's performance if we made it less than 3.
Dean Armbruster
System Analyst
757-989-2839
-Original Message-
From: u2-us
If processes are putting things in a queue for a phantom to process and
you use PAUSE, the processes can use WAKE to signal the phantom that it
has something to do. This way, the cycle time can be much longer,
reducing the number of periodic checks. The result is better response
time on items tha
Dan, what version of UniData are you running. I reported a similar bug
several years ago and it was fixed in version 6.1.7. With that bug, it
happened when the process was killed, and then even logging out wasn't
sufficient; any process reusing the UDTNO of the killed process could no
longer PAUS
FILEINFO
25 - Returns update trigger
26 - Returns delete trigger
Dean Armbruster
Information Technology, System Analyst
Ferguson Enterprises, Inc.
12500 Jefferson Avenue * Newport News * VA * 23602-4314
T: +1 (757) 989 2839 * F: +1 (757) 989 2801
www.ferguson.com
-Original Message-
Fro
I also find it interesting that it isn't in the UniBasic Commands
Reference manual, even though 27 & 28 are there.
We saw 25 & 26 in a README for a previous release and have been using it
ever since.
Dean Armbruster
System Analyst
757-989-2839
-Original Message-
From: u2-users-boun...@l
>From BASR.pdf, UD 7.2:
27 Is the file encrypted? ALL 0 - File is not encrypted 1 - File is
encrypted
28 Type of file encryption ALL Returns a dynamic array containing the
following information:
? For a file encrypted with the WHOLERECORD option:
-1@SM@SM
? For a file encrypted at the field lev
Often, when we use TANDEM to fix a udt that is in a do-forever or
otherwise "stuck" (in the users perspective), sending a BREAK (ESC+D)
does not work and we can't break in to the BP. We have discovered that
the reason for this is that the user, in their attempt to free up the
session, has entered
I am not certain about AIX, we run on hpux, but I believe unirpcd can be
restarted on AIX also. We do it occasionally for just the symptoms you
describe a few times a year.
>From a system prompt >
>stopunirpcd
>startunirpcd
The rest of UniData keeps on running fine. No disruption.
Dean Armbru
@DATE and DATE() are not the same thing. Neither are @TIME and TIME(),
as demonstrated by Colin's test below.
>From the UniData Commands Reference document:
DATE() - Function returns the current system date in internal format.
@DATE - System date when the program began in internal format.
Since
Actually, that is the UniBasic Commands Reference, not UniData Commands
Reference.
-
@DATE and DATE() are not the same thing. Neither are @TIME and TIME(),
as demonstrated by Colin's test below.
>From the UniData Commands Reference document:
DATE() - Fu
Were you using the OVERFLOW option with memresize? If not, memresize
should not be creating the extra over files. If memresize did create
the extra over files without the OVERFLOW option, then that would be a
bug in memresize.
Dean Armbruster
System Analyst
757-989-2839
-Original Message--
The ID of _PH_ entry is not Z_. It is Z_.
is 3 digits, zero filled. is not zero filled.
Note that it is capturing level, not EXECUTE level (@LEVEL). If you have
4 levels of EXECUTE, but only 2 have CAPTURING clauses, in the lowest
child @LEVEL will be 3 but the capturing level will be 001 in t
Are you certain that the dbpause finished? We've had cases here where
our backup scripts would hang on the dbpause command. Those cases were
not random however. They all happened after a "file system full" event.
When a file system is full and a UniData write occurs that results in a
file being
Is there a way to programmatically determine the number of licensed users
allowed?
SYSTEM(514) reports the number of non-phantom users currently logged in, which
is essentially the number of licenses in use. I want to compare that against
the number of users allowed in a monitoring program. I
And the winner is:
PORT.STATUS
It is quick and easy. My program executes it using its own @USERNO and parses
it out. With no options, it returns instantly.
LISTUSER and listuser were too slow. It takes 20+ seconds for them to return
from an captured EXECUTE/PCPERFORM when it has to output o
What is the difference between LISTUSER USRTYPEs "phantom" and "iphantm"?
On our systems, we will have 100 to 200 phantom processes, i.e. background
processes, running at any given moment. When I look at them via LISTUSER, they
will be split approximately 60/40 between phantom and iphantm. All
Have you used fileview or even vi at the unix level to look at what is
in the file?
We had a problem like that back when we were on 6.1.18 (hp-ux 11.11).
When we viewed the corrupt groups in the file with fileview, what we saw
was output from a process, as it would appear if using a CAPTURING
clau
Is there any way to expand the number of items stored in a SAVEDLISTS
list?
We've hit our system's limit on savedlists size. The limits of 1000
sublist files and 34,815 bytes per sublist are not enough to hold some
of our savedlists. Is there any way to expand the sublist file range
from 3 to 4
I did make a bad assumption about a savedlist being limited to 1000
parts, but there is a problem that I'll have to submit a case for.
Since the parts default to a 3 digit numeric suffix but will add more
digits if more than 1000 parts are needed, the possibility exists for
another list to stomp o
If you're not yet up to 7.1.15 or higher, you can fake df so that smm
will not hang.
Like this:
1. Rename "/usr/bin/df" to "/usr/bin/df.real".
2. Create a script to replace "/usr/bin/df".
#!/bin/sh
cat /tmp/dfoutput
3. Create a cron job to occasionally update the cat'd file using the
While it is true that Ferguson is on a path to SAP, the rollout is still
several years out. Until then, we are still enhancing our current
UniData applications. With the rollout dates still uncertain, and the
off chance that the change will be cancelled, we aren't about to leave
our user base fal
My apologies for the poor job I did posting this. I sent it before I
had completed the subscribe procedures. I blame it on my vacation. :-)
Because of my mistake, I didn't get the replies. I had to get them from
the list archive. Here are my replies to the replies:
To Dave Davis, who said,
SELECT with the SAVING keyword, with or without NO.NULLS, returned the
correct value in @SYSTEM.RETURN.CODE, in my tests.
I don't know or have Universe, so I could be wrong, but it appears you
have an implied BSELECT, as demonstrated with the SAVING keyword. A
BSELECT is, as quoted from the UniDa
BY.EXP does not factor in. This is about BSELECT, regardless of BY.EXP
use.
@SYSTEM.RETURN.CODE is not supposed to be the number of *records*
selected. It is supposed to be the number of items in the select list.
>From the UniData UniBasic Commands Reference (BASR.pdf), Appendix B:
ECL @SYSTE
According to the UniData documentation (UniData UniBasic Commands
Reference (BASR.pdf), Appendix B), @SYSTEM.RETURN.CODE is supposed to be
the number of items in the select list. This is for all the various
SELECT commands, (SELECT, SSELECT, NSELECT, QSELECT), not just BSELECT.
Using the example
I didn't see a post for B, so here is mine:
Inside your trigger routine, use SYSTEM(49) to dump the program stack.
@LOGNAME will give you the user running it. Report them in a fashion
most helpful to you, e.g. write it to a log, send an email, etc.
SYSTEM(51) may also be helpful.
> -Origin
What causes an "Expresion stack underflow"? (Error #40198)
A couple of our developers here are getting this message:
In at line Expresion stack underflow.
I have unconfirmed reports that it happens occasionally when the main
program or subroutine is recompiled with minor changes while in
I'll be there.
> -Original Message-
> From: Jerry Banker
> Sent: Wednesday, August 08, 2007 1:07 PM
> To: u2-users@listserver.u2ug.org
> Subject: [U2] U2 University
>
> I'm going to the U2 University in Denver the last week of
> this month, any others attending?
>
>
>
> Jerry
---
> From: Barry Rogen
>
> Question: We are currently very active with EDI documents with
>most of our major accounts. We have been looking into
> starting the '820' document along EFT (elec.
> fund trans).
> Has anybody been doing the 820 along with
Does anyone know what "UPL File lock error (105)" means?
We received this error in a UniBasic program on a CLEARFILE command on a
temporary file created at the start of the program. The file has no
indexes.
I did RTFM, but could not find any references to it in. The list
archives mention UPL er
It's been my experience that the main benefit to a degree is getting the
resume past the Human Resources department. HR personnel are generally
not technically oriented and therefore are screening for candidates
based on set criteria, i.e. degrees and specific buzzwords. For a
non-degreed candida
While quibbling over nanoseconds, some of may have missed a little flaw
in your tests that would have nothing to do with caching or memory
allocation. The elapsed time of each test could have been different
enough that the differences in the counts would fall within that margin
of error. This is
The time period of the test does not matter, nor the CPU speed. The
results for each programming style were still close enough in each test
to fall within the margin of error of a partial second.
My apologies for rekindling the holy war. That was not my intent. I
was trying to explain why the s
With a larger buffer, memresize can shuffle the data around in memory,
then write it the new file. This saves I/O because it can write the
file out in whole groups instead of rereading a group to insert each
record as it runs through the original file.
If the file is larger than the buffer, all t
Is there a way, short of "EXECUTE UDT.OPTIONS CAPTURING STUFF" and
parsing STUFF, to determine the current setting of a particular UDT
option?
We're running a program where we need an option set a certain way, but
we want to be able to set it back to it's original setting before
returning to the c
I just used the 'stopsmm -r' command with version 6.1.10 on HP-UX and it
worked as advertised for both a number of users licensed update and for
an Authorization update. It certainly is nice to avoid the headaches a
restart.
My only comment is that if this handy way to bypass a restart has been
a
Any UniData Journaling users out there?
We are considering using journaling as a way to keep a near real-time
backup of our primary system. We would be journaling hundreds of files
on a system with over 1000 users.
We have lots of questions, like:
Has anyone used journaling on that scale before?
My apologies. I forgot to post what we're currently running: UniData
5.2.28-64 on HP-UX 11.11.
We're currently on an upgrade path to take us to 6.1.10 in the near
future. Since DR was new with 6.0, we'll now be able to take advantage
of the DR capabilities as you suggested. We're looking into
The RETURNING clause returns a multi-attributed array of error codes, one
attribute for each error. The attributes themselves may be multivalued,
depending on the error. The first value is the error code, as found in
ERRMSG/ENGLISH.MSG. Any other values are data to be inserted in the error
m
>From within a UniBasic program, is there a way to determine what started
it, i.e. a proc, a paragraph, or another UniBasic pgm?
UniData 5.2 and 6.1
HP-UX
Long story:
When using COPY with a select list active, COPY will prompt "get id's
from select list:(Y/N)?". The exception is: COPY will not
p
That is what we are trying to do. In order to ensure that we don't
break any existing programs or procs using COPY, our replacement needs
to emulate the real COPY in all ways. We've covered all its uses
(with/without select lists, to same/different files, pick/UniData mode,
overwriting, deleting,
@SENTENCE is always the name of the program we're in.
@COMMAND varies, but it is not consistent with the appearance of the
prompt.
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] On Behalf Of
> [EMAIL PROTECTED]
> Sent: Friday, July 29, 2005 11:56 AM
> To: u2
The trick, from my original email:
When using COPY with a select list active, COPY will prompt "get id's
from select list:(Y/N)?". The exception is: COPY will not
prompt if executed from a proc.
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] On Behalf Of
>
No udt option that I can see. It's one of the first things we checked.
I just now noticed that I wasn't clear enough in my original post. It
happens when copying to a different file or item, not when copying to
the screen (or COMO). Which is one more darn thing we'll have to code
around. (sigh
When a CAPTURING clause is used, UniData creates a temporary file in the
_PH_ file to hold the captured text. The format of the file ID is
'Z_nnnp' where nnn is the capturing level and p is the process
ID (@USERNO). P is not zero filled, i.e. it could be 1 to 5 digits.
Note that the
According to the 4.1 Administering UniServer doc, ODBCLOGIN can be a
globally cataloged subroutine, not a login paragraph.
It's funny how there is no mention of this in the 5.2 or 6.1 docs. But
then, there is no reference to UniServer either, except for a few vague
mentions for environment variab
PCPERFORM can do multiple commands. Separate them with a linefeed
(char(10)). You can actually pass in a text stream that is a script for
the shell if you like, with if's and else's and other stuff, with a
linefeed between lines. We use this run commands in other accounts
without leaving the cur
Add a process to the LOGIN paragraph that writes information about that
udt process, such as it's path, to a file. That file can be queried to
determine all paths.
Don't forget to add something to the LOGOUT paragraph to remove that
information so that the file is always current.
> -Origin
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] Behalf Of
> [EMAIL PROTECTED]
> Sent: Thursday, May 06, 2004 9:26 AM
> To: [EMAIL PROTECTED]
> Subject: [U2] Is there a better way?
>
>
> I've got an attribute with value marks and sub-value marks.
>
> Is there a
I have a life, and I enjoy some good humor. I also appreciate it when the sender
changes the subject line when the topic steers off course, and adds [OT] when
appropriate. Some seem to have forgotten those courtesies.
Thank you, and have a great day.
Dean
> -Original Message-
>>
>
> In a message dated 4/30/2004 2:07:27 PM Eastern Daylight
> Time, [EMAIL PROTECTED] writes:
>
> > What we have is a case where an open file error may occur
> and we need to
> > have the user go back into a menu, but we do not want them
> to see the
> > error. Problem is that the error is wai
> On Behalf Of Dana Baron
> Sent: Monday, May 03, 2004 2:59 PM
> To: U2-Users
> Subject: RE: [U2] PHANTOM forever
>
> I have a Unidata Basic program (called MESSAGE_LOOP) that needs to
> constantly check for messages deposited into a particular
> directory. It is
> an endless loop that checks for
53 matches
Mail list logo