Rocket MultiValue

All,      So we have been battling some extreme system slowness intermittently for years. When this comes up we max out on LCT's, the AIX kernel goes into a panic causing thousands of processes to back up in CPU wait. In  our case it was always nfauser.       We were able to eventually track it back to some infrastructure we had implemented for opening files. Some of the code was basically opening up an NFA  file pointer for each record selected in a loop, so if it was thousands you got thousands of opens and thousands of remote nfauser sessions. We found it and changed it to open once and seemed to fix it for a while. Then it was still happening intermittently. With the previously mentioned bug we saw this on 3 of 4 production systems, once fixed only on our heaviest nfa accessed data server.    What we were able to determine with help from Rocket support was that upon log off UniData spawns a processes that scans /usr/ud/tmp, savedlists, and SAVEDLISTS to cleanup any transient work f

We are running 2 versions of U2WebDE (AKA Redback) 4.4.3 5.3.0 on our Linux box.  Both versions have 3 environments that connect to our UniData accounts: DEV - used by programmers (me) to write new stuff PILOT - used by super-users to test changes prior to loading to LIVE LIVE - available to anyone who can access the web pages   We have been smoothly running this way from over a year with the plan to phase out the old version (4.4.3) by the end of the year.Yesterday, a web page in our PILOT (user testing) environment started to fail - hard!  The lines of the web page making the call are as follows: accnt = Application("ERPAccount") userid = Application("Ruser") passwd = Application("Rpass") set oOrder=CreateObject("RedBack.RedObject") oOrder.Open2 accnt,"SF:oOrder",userid,passwd The oOrder.Open2 is failing hard enough that I have to reboot IIS to get control back. I see no errors in the IIS logs or the Redback logs. Nobody has changed anything in the web pages or back end logic (it

I have the personal edition unidata 8.2.4.3001 and it looks like the openssl version is  3.0.1 can someone confirm to me that this is the same version packaged with 8.2.4.xxxx official? thanks  dougc------------------------------Doug ChancoUniverse ProgrammerSelf Registeredgraham NC US------------------------------

I am using a install.parmas file with the following: (modified from the default provided) LicAgree        YesStopSys         YesUdtHome         /usr/ud82UdtBin          /usr/ud82/nomUpgrade         YesWorkDir         /usr/ud82/workLibDir          /usr/ud82/libLogDir          /usr/ud82/logUnisharedDir    /usr/ud82/unisharedGroup           sysSerial          LinuxInstallXDEMO    NoRFSflag         YesREPflag         YesREPsysid        dendevmvasbld11_82Backup          NoConfProduct     NoPreservPypath   YesPreservPymodule YesOverWrite       NoConfirm         YesTransMEK        NoSaveDir         saveOldBIN          .CheckLang       YesCheckPerms      YesStartud         YesFixPerms        NoCheckPerms      Yes when I run udtsetup -f install.param I am getting this error: Welcome to the Rocket UniData RDBMS. http://www.rocketsoftware.com/u2 Using /ud8243001/bin.tar during this process. License files are located in $UDTBIN/license. Parent directory of /disk1/agent/workspace/ud_build/src/udh

Hard drive backups that I try to FTP to a cloud server never finish. Here's the tape line from the pick0 file: tape /compass/picksave 500 f lx  # tape 3. I know the complete file has been uploaded because if I close the FTP program (and I've used four different ones) the file is usable from the cloud for a FULL RESTORE. Below is the part of the FILE-SAVE that writes out the file to the Linux hard drive. There's a T-DET at the end of it. Any idea why the ftp protocols don't know that the upload / download is complete? Thanks and a tip of the hat. HT-SELECT 3PHT-REWPHIF E = 93 XOO FILE-SAVE Beginning at +HTIMEPHSAVE SYSTEM (TSFSTONH<PHT-WEOFPHOO Dumping STAT-FILE to Tape...HT-DUMP STAT-FILE (I)PHT-WEOFPHHT-REWPH ------------------------------Richard GinsburgPresidentGinsburg ConsultingManassas VA US------------------------------

Gedday everyone, We have MVIS set up with subroutine and data resource end points for various data retrieval. There was a recent request from the business to be able to supply a list of keys in a request to retrieve one or two specified fields in the response. The number of keys was in the order of 20k-40k. So I did a test and squirted a request (using curl) to our dev MVIS containing of 10k keys. It timed out. I then tried the same request with 5k keys and after a few seconds a valid response was received. Here is the url used. http://mvis.mydomain.com:7171/TEST/DATARESOURCE?select=key=%22KEY1%22%22KEY2%22%22KEYn%22&fields=FIELD1%2Cxref%2CFIELD2 Next we decided to just use the built in limiter (max) via a similar request and found that we could obtain the desired volume in the response. http://mvis.mydomain.com:7171/TEST/DATARESOURCE?max=20000&fields=FIELD1%2Cxref%2CFIELD2 Is anyone using (or has played with) large queries in MVIS. Is there a limit in MVIS or a setting (MVIS/U

Hi Guys, Other than the rp_driverux.c example, is there any documentation regarding writing OSFI drivers? I have a need to write one to update a postgresql database.  I have it working, except for two issues: I cannot figure out the protocol for RP_GETGRP when an item exceeds the 4000 byte buffer.  At present I'm using the ux.c method of returning RP_CMDSUP to make the server backtrack to using RP_SEQIDS, but I can see certain efficiencies are lost when the system now has to make a RP_RETIX call for each item in the list when it would otherwise have the record data available directly from the request for the next group.  I have RP_GETGRP working when I can fit each group of id's and records in the buffer, but it fails with forward link zero errors when the group is larger and won't fit. The second issue is more pressing and might be a deal breaker for me.  Is it possible to call a basic callx trigger when writing to a file served by an OSFI host? Thanks Tom ------------------------

We have a strange situation. One of our applications written in vb.net has it's data etc hosted in Azure.  When running the application from the office over a site to site vpn to the Azure endpoint everything works perfectly well eg a typical report runs in about 5 seconds.  However, when the users connect over the Azure software vpn on their home machines (all users regardless of local internet performance) the whole application slows down dramatically eg 90 seconds plus for the same report.  The IT Infrastructure team have exhausted their investigations and can find not reason why. Running commands at TCL using a Telnet application all seems OK but when running the app (uses port 31438 for RPC) is when the speed issue is experienced. Has anyone seen this or have any suggestions on where we might need to look?

I am being asked to break a tab delimited file as part of an assessment. How do I identify the tab delimiter?------------------------------Kimberly CrumRocket Forum Shared Account------------------------------

Hi, We're working on an archive process and we set it up to run as a phantom. When run as a phantom, the process runs over twice as long as it does on a telnet session. Does anyone know a why to bump up the priority of a phantom job on a Windows server?  Thanks Curt------------------------------CURT VALENTINESOFTWARE ENGINEER IIIPaychex------------------------------

we mainly use mvis with a subroutine, see definition below.  currently, if the input JSON has the wrong input data element we return a JSON array with error string. however, when use this the HTTP return code is always 200. this means other systems have to look for error strings in response. is there a way to raise the HTTP error code to something else other than 200  to something like an error code of 400 to indicate an error occurred? 

I'm trying to create something in linux that I can run from the accuterm esc:stx command. I wanted to connect to FTP, login, and send a get and/or  push command but it doesn't seem to be executing the statements: ftp nnn.nnn.nn.nnnibmftp**passwd**cd /BizApp/Parts/Testbinaryput "KAlside.pdf" bye I've tried 'echo' and using the <  to pass in the data... not working. Anyone have something like this? Thanks------------------------------Kathleen HambrickProgrammer at Colwell------------------------------

There are numerous approaches to turning output pagination OFF from within BASIC code (and I'm not really sure which ones work and which are just hang-overs from other MVDB implementations). What I want to ensure is that output pagination is turned back ON from within BASIC. Does anyone know if a method for achieving this? Thanks ------------------------------Gregor ScottSoftware ArchitectPentana Solutions Pty LtdMount Waverley VIC AU------------------------------

I'm working on something to make deployments to new environments a little easier, and one thing that would be super helpful is the ability to create or update VOC items from a script of some kind (e.g. a PAragraph).  I could of course do it from a basic program, but then the person doing the process would have to know how to save the program to the appropriate place, compile and catalog, and then run the program.  The idea is to give them a script and after saving it as a PA in the VOC, have them run it and it will take care of all the heavy lifting after that.  Most things I can do (creating directories, compiling programs, etc) but can't find a command that will allow you to update a file directly from TCL.  Anyone got any suggestions on a command?  I've got to be missing something obvious that will do this. PA UPDATE VOC INVOICE.EXTRACTS SET INV.PATH='/path/to/invoices' UPDATE VOC INVOICE.EXTRACTS SET INV.PATH2='/path/to/invoices2' UPDATE VOC INVOICE.EXTRACT

We are experiencing a large number of odd jbase_agent processes being spawned that are interfering with system operation (users experiencing freezing, hangs, unresponsiveness).  The processes use command options I don't recognize. Our normal jbase_agent is running with these processes as per ps -ef | grep jagent: jagent     3394      1  0  2022 ?        00:00:00 /bin/sh /home/jagent/jagent_20002.shjagent     3556   3394  0  2022 ?        00:25:48 jbase_agent --config=/home/jagent/jagent_20002_config Meanwhile, the spurious processes look like this: jagent    79624      1  0 12:47 ?        00:00:01 jbase_agent -w --config=/home/jagent/jagent_20002_config +A 5 +H 8 I was unable to find documentation for a lowercase -w option for jbase_agent; I also have no idea what the "+A 5 +H 8" at the end of these means. Can anyone help? Apologies if the forum is not the right place for this question: I tried creating a proper support issue initially, but when looking for a jBASE support portal I

Hi, I have the following problem: I have been using the Rocket U2 product (Universe|Unidata) as ODBC configured on a SQL Server 2014 Via Linked Server Tecnology and everything It works correctly, I connect the U2 against my Universe database and this with the ODBC of Universe for Windows Server, finally I use the ODBC to create the Linked Server in SQL SERVER 2014. The problems began when upgrading SQL Server to SQL Server 2019, from that moment with the same configuration the linked server has stopped working. Have you reported any ODBC bug of Universe in SQL Server 2019? Does anyone have a similar scenario? I have tried several configurations both in U2 (via Uci Editor) and in the SQL Server and nothing has worked (The linked Server is infinitely loading and dont retrieve data from U2 database).            

Thanks Brian, and yeah, I've used the SQL functions over ODBC before, butit's kind of fiddly to setup. The intent here is to provide something tobootstrap an inexperienced user in installing some new code and artifactsthat go along with the code. You can type "SQL UPDATE ..." but only afterhaving setup the account for SQL (I think?). I get "No privilege on fileVOC!" as an error at least.I was trying to get away from having a basic program, because it'sobviously challenging to a non programmer to understand all the possibleways compiling and cataloging a program can fail. But installing a utilitylike UPDATE.RECORD once and cataloging globally might be the bestcompromise option. I can write my own, with a pretty rigid syntax like"UPDATE.RECORD VOC TEST SET 3=nk jhkjewqh ekjhqwkje jk" since it's notdesigned to be used interactively, just used for updating parameters wheninstalling some new code.Really appreciate the suggestions!On Mon, Feb 27, 2023 at 11:49 PM Brian Speirs via Rocket

Hi all, Happy Monday Morning ;) I am looking for help with an AWForm. Hopefully some of you are pretty good at this interface. I did look through what I have as documentation and did not see this information listed but - we undoubtedly have old and incomplete docs. We have quite a few awforms in use, and I see examples of columns in many. What I am trying to do however is to load the data into a table format such that excel would use. Namely we want the lines between the MV data vertically and horizontally. While a tab will move between cells in the created form, passing in a tab at the end of the first line did not separate the data from MV1 to MV2. There is a horizontal line at the bottom of the array, but not between each line of data. Suggestions?? Here is what my form looks like: PRESS_TYPE CUSTOMER JOB # Due Date Notes CNO JN DTE       PART NUMBER CHIP SIZE RADIUS MATERIAL RM # COLOR NAME HOLE SIZE DESCRIPTION/NOTES RM DIMENSION PROFIL

I have been converting code from R83 to Prime Information, Advance Pick to Unidata, Sequoia Pick to Unidata, D3 to Universe, and even Unidata to Universe, I have seen some really awful code and very morbid. Of course, you are thinking some of the latest Basic commands that I use daily in Universe did not exist in R83 so those programmers did the best they could. While I do not disagree, I am looking at code from customers in 2023 that should have been updated sometime in the last 20 years or at least in the last 10 years. I have detailed below my comments with what I found and what I think it should be. I will ignore for the time being GOTO, GOSUB, and ON RESPONSE, for example. I would love to see what you have found as well because it just makes me LOL when I see it. * Using COUNT instead of DCOUNT MY.ARRAY = 5:@FM:7:@FM:8 MAX.FM = COUNT(MY.ARRAY,@FM)+ 1 * Should be MAX.FM = DCOUNT(MY.ARRAY,@FM) * Not knowing about WHILE in a LOOP statement EXECUTE 'SELECT VOC'

while we suspect a 3rd party program might be causing the issue or at least helping it along , what we are trying to do is isolate or prove it. any suggestion or ideas are welcome as we are running out system info udt - 8.1.2 redhat - 7.9 when unidata locks up  showud shows all the processes running yet any ECL sessions are locked up trying to login via udt hangs, we turned on error logging but the log file only shows the start of unidata and nothing else udtmon - does show a little disk activity we did straces on the udt processes including trying to open a udt session  attached is an strace of a udt session and an ipcs output when the system is hung any ideas or suggestions are welcome as well have been stumped, we are talking to the 3rd vendor as well and like I said we are not positive that the issue but we have seen their process using a lot of memory and cpu stopping unidata and starting it back up and everything works fine, for a while, there is no pattern to when it lock

I am planning upgrading our D3Linux system and the hardware it runs on.  I am considering setting it up to run on a RHEL virtual machine with the hope that future major version upgrades might be easier on the same hardware. Are there any major cons of running in a VM?    Any configuration and setup tips are welcome.------------------------------Steve CaspersMurphy Automotive IncRosemount MN US------------------------------

Hi All, I'm sure I've posted on this matter before but I just had a bit of a wobble understanding why my code wouldn't compile, until I notice the variable I had used started with the word NEXT (as in Next_Available) and the D3 compile went out and took another shot of tequila and said 'nada - not today hombre'So, just put it out there NEXT (suffix not allowed) and I'm sure there may be others - like NOT and FOR and ...Does anyone care to comment?------------------------------Stefano GallottaManaging MemberSimply Red Open SystemsMilnerton ZA------------------------------

Just noticed that Unidata had a directory called minidump When i looked at files in the folder the files stated "The thread tried to read from or write to a virtual address for which it does not have the appropriate access" What does this mean, should I be concerned, how can I tell what is causing these files to be created------------------------------robert modrichDeveloperRocket Forum Shared AccountTEMECULA CA US------------------------------

Hi Folks, we recently upgraded to 82 Unidata in a testing environment, and are getting the following in SMM logs; In UT.OBJ/_S.READLIST at line 90 insufficient memory, requested size = 37567064. Thu Feb 16 14:28:02; errno=0: ICSubCall returned from basic, ReturnCode=-1. I do not have access to the source code for S.READLIST but it is dealing with very large save lists. Any thoughts on what we might have wrong? Rick Brown------------------------------Rick BrownSysadminRocket Forum Shared Account------------------------------

Anyone using UniVerse and EDA with "virtual" dictionaries?  I have a user running UniVerse 11 and using EDA. EDA is working fine on true attribute dictionaries, but now he's trying to add an I-type like this:      DIV0001 I0002 "10"00030004 DIV0005 10R  Just a hard coded value "10" - but in the SQL Table the column is populating with NULL. ------------------------------Mark EastwoodDeveloperAFS Technologies IncPhoenix AZ US------------------------------