Rocket® U2 | UniVerse & UniData®

The default uvconfig values for these settings are:N_AFT 5000 N_AFT_SECTION 17 N_AFT_SECTION_BUCKET 23  These result in 294.12 files per SECTION, and 12.79 files per SECTION BUCKET. The N_AFT setting is too small for our needs so we will be increasing it. What I am interested to explore is what are the deciding factors needing consideration when adjusting N_AFT_SECTION and N_AFT_SECTION_BUCKET ? Should the current ratios be maintained as best as possible? For example, if N_AFT is set to 15000 a custom setting of 53 for N_AFT_SECTION would result in 283.02 files per SECTION and 12.31 files per SECTION BUCKET, which are close to the default ratios.Is this the best approach, or should I take a different track? 

Docs on this setting describe it thus:Sets the number of hash buckets in the per-user locking table. Adjust this based on the number of concurrent locks a user may hold.andThe number of hash buckets for the per-process locking table. This parameter is highly application dependent, If the application requires a large number of locks in one transaction (more than 20), you should increase this setting to the closest prime number to the maximum number of locks per transaction.  So if we have identified our maximum locks per transaction is a rather large number, like 5000, then this calculation would derive to 4999. That does not seem to make sense.Is there any real-world feedback on this?

We are running Universe 11.3.1 on AIX 7.2.  I have 9 production accounts, mirrored on development and test LPARs.  That all adds up to thousands of files.     Formerly we used Fast c/s to monitor the entire system and stay on top of undersized files to prevent application slowdowns.  Unfortunately, it is no longer supported.Does anyone know of a similar application or product we could use?  If so, could you speak to the efficiency, ease of use etc.?Thanks in advance.

Greetings Rocketeers!We are on WIndows, preparing to move from Universe 11.xxx to 14.xxxHave any of y’all made this upgrade recently?  Were there any functional issues after upgrading?  Is there a test plan available anywhere we could get our hands? Many THX,

Greetings all!In UniVerse 11, how are the 7 records associated with a port file created? >SELECT 405P7 record(s) selected to SELECT list #0.>>LIST 405PLIST 405P 01:57:21pm  30 Sep 2026  PAGE    103P.......ONLOCMNULOGPONTFDC

We are having trouble connecting via unirpc (MVQuery) to our Unidata 8.4 (Windows 25) server.  My work PC does not have a problem.  While RDP’d into the server, when I try to connect to an account, I’m getting → So far, 2 other PC’s get the same message.  I’m trying to setup logging/tracing for this.  I’ve placed ...\include\ud_database into TRACE_LEVEL=2.  If this is producing output, I can not find it.  What else can I do to track this down? Thank you,Jonathan WellsAptron, Corp.

Welcome BackHi, MultiValue Community!The second half of 2026 is packed with opportunities to learn, connect, and plan for the future of your MultiValue environment. From roadmap insights and AI innovation to community events and modernization resources, we're focused on helping you maximize the value of your existing investment while preparing for what's next.This month, we're highlighting upcoming educational opportunities, customer events, product direction, and ways to engage with fellow MultiValue customers around the world.Featured Event: MultiValue Roadmap Webinar—Get a Preview of What's AheadWatch the recent MultiValue Roadmap Webinar to hear directly from product leadership about future priorities, innovation investments, and customer-driven enhancements.Topics we covered:Product roadmap updates AI opportunities across Rocket solutions Modernization initiatives Customer-requested enhancements Upgrade and migration planning guidance Q&A with product expertsWatch the recordin

HiSBClient SBXA 6.6.2. Character mode, not GUI. We have our own menu file (CSMENUS) where we /MD to create menus with processes & programs as the menu options.For each menu option listed within the menu, I would like to add some F1 help text for the user to see when using the menu & prtessing F1 on each option.But I cannot figure out anywhere to input this or if it is possible. When pressing F1 on a menu option within a menu now I get [{MENUNAME}.REM,{POSITION}] NO HELP MESSAGE AVAILABLE.Is it possible to get to this & enter some help text?Full example:Menu file CSMENUSMenu name GE (General enquiry)/MD SetupMessage when using menu & pressing F1 currently.  I’ve searched the forum & documentation but can’t find anything.ThanksJack

Missed This Week's MultiValue Roadmap Webinar? Watch the Recording on Demand.Thank you to everyone who joined our recent Innovation, AI, and the Future of MultiValue webinar sessions September 15-16. We had fantastic participation from the MultiValue community, along with excellent questions and discussion around product innovation, AI, roadmap priorities, and upcoming opportunities for customers. If you were unable to attend, or would like to revisit the content, you can now watch the recording on demand:👉 Watch the Webinar Recording:Innovation, AI, and the Future of MultiValue.During the session, we covered:✅ MultiValue AI insights and roadmap direction✅ Upcoming MultiValue beta programs and release highlights✅ Product roadmap updates across the MultiValue portfolio✅ The Rocket MultiValue Accredited Solutioning Partner Program✅ eLearning resources, documentation, and community initiatives✅ The MultiValue Innovation Library and customer engagement opportunitiesWe also received valuab

There’s a lot happening across the Rocket® MultiValue portfolio, from AI initiatives and upcoming betas to roadmap investments, learning resources, and community programs.If staying ahead of product innovation is important to your organization, now is the time to register.Join our upcoming Innovation, AI, and the Future of MultiValue webinar to hear directly from the MultiValue product management team and get a firsthand look at what's coming next.The sessions are free to attend, and all registrants will receive access to the recording.Reserve your spot today:Session 1 - Sept. 15, 4 p.m. MT (Denver) | Sept. 16, 10 a.m. UTC+10 (Sydney) Session 2 - Sept. 16, 10 a.m. ET (New York) | 3 p.m. UTC+1 (London)During the webinar, we'll cover:✅ MultiValue AI Insights, upcoming beta opportunities, and AI roadmap updates✅ 2026 product betas, roadmaps, and release highlights✅ Product versioning and lifecycle updates✅ The Rocket MultiValue Accredited Solutioning Partner Program✅ eLearning and informa

Hi everyone, I’m attempting to install the UniVerse Personal Edition version 14.1.1 on MacOS Tahoe 26.1 on an M1 Macbook. I’m following Rocket’s official PDF (see attachment) on how to do this via a Docker container and using the supporting files from their Github: RocketSoftware/multivalue-containers Because I’m on Apple Silicon, I adjusted the Dockerfile as follows:Replaced the original FROM to force AMD64 emulation since I’m on ARM:FROM --platform=linux/amd64 rockylinux/rockylinux:latestI also added this to fix a permission error that occurs.RUN chmod +x /usr/local/bin/*uv.sh Now, I consistently hit a segmentation fault in the installer during the init_uvsm step. Below is an excerpt of the output logs where the error occurs after running docker compose up:universe-server  | universe-server  | Starting UniVerse...universe-server  | UVSMM is started.universe-server  | Starting uvsm...universe-server  | /usr/uv/bin/uv.rc: line 562:  3654 Segmentation fault      $uvbin/init_uvsm $init_u

Getting the below issue operating replication with EDA.Hoping someone can shed some light on how to resolve our issue below.U2\BIN\RW.ERRLOG u2\bin\rw.errlogEDA SQL Server Driver: [Microsoft][ODBC Driver 18 for SQL Server]TCP Provider: An existing connection was forcibly closed by the remote host.Mon Aug 24 06:20:53 RW(0,7460) report:Mon Aug 24 06:20:53 RW(0,7460) report:In E:\U2\sys\CTLG\c\CXG.WRITE.EDA.BI.DATA at line 404 Mon Aug 24 06:20:53 RW(0,7460) report:EDA_write_tuple error, id = 'A143088'Mon Aug 24 06:20:53 RW(0,7460) report:The EDA file is CXSFILE stored on the Unidata Subsciber. We get the above error attempting to write to CXSFILE (EDA file) with aTrigger on CXFILE as below configuration.CREATE.TRIGGER CXFILE CXG.WRITE.EDA.BI.DATA AFTER UPDATECREATE.TRIGGER CXFILE CXG.WRITE.EDA.BI.DATA AFTER DELETESome evidence supports: client-side dead-handle defect, stale held connections are not detected and never re-established.I believe we have covered off root cause with Subscriber

The Rocket Software MultiValue team is pleased to invite all Rocket UniVerse customers with an active maintenance agreement to participate in the UniVerse 14.3.1 Beta Program.This is your opportunity to gain early access to upcoming enhancements, evaluate new capabilities in your own environment, and provide direct feedback that helps shape the final release before general availability.OpenTelemetry Support (Phase 1)Gain greater visibility into system behaviour and performance through the first phase of OpenTelemetry integration, providing a modern foundation for monitoring and diagnostics.Enhanced Auditing CapabilitiesThis release delivers a number of improvements to UniVerse auditing capabilities, including:User-defined audit event handlers Enhanced audit configuration & logging General audit usability and reliability improvementsUpgradesPython 3.14 OpenSSL 3.5Beta Program DetailsBeta begins: September 23, 2026 Beta concludes: November 4, 2026 Supported platforms: AIX, Linux, and

I've been building a tool that puts a MultiValue account under real version control. It runs on UniData and UniVerse today, and it's built to port to any MV system — the engine sits on a deliberately narrow set of record primitives, so a new platform means implementing that handful, not the tool. jBASE is in active development — its test suite is passing and I expect it in 2.1.What it does. It commits the account's own records directly to git objects — your records, your dictionaries, your VOC. The live account is the working tree, so there's no export copy drifting out of step and no import step to forget. Change a record, GIT STATUS shows it changed.Both surfaces work. From the shell it's udt-git / uv-git, taking the same subcommands and flags as stock git. From inside a session it's a cataloged GIT verb — GIT ADD -A, GIT COMMIT -m "...". Since 2.0.0 both commit identical bytes, which took longer to get right than everything else put together: a dictionary isn't stored the way it's h

Rocket® MultiValue Pulse | August 2026Welcome BackHi, MultiValue Community!Welcome to Rocket® MultiValue Pulse, the refreshed newsletter experience for the MultiValue community—created to bring you timely customer updates, practical resources, and community conversations in one place.As we move through the second half of 2026, our focus remains simple: helping you maximize the value of your MultiValue investment while preparing for what's next.For many MultiValue organizations, that also means ensuring critical applications remain stable, supported, and ready for the future. In upcoming editions, we’ll share resources and conversations focused on business continuity, long-term planning, and practical steps to help teams reduce risk while continuing to modernize.Over the coming months, you'll hear more from us through customer webinars, community discussions, product updates, and practical resources designed to help you strengthen operational resilience, modernize with confidence, and c

We have may programs that use basic format like “R%8” to prefill zeroes that were not entered or came from Excel who loves to strip them off.We just updated our development server from 11.3.1.6022 to 14.2.1.If your basic program takes a number in missing leading zeroes, “R%8” would turn 1234 into 00001234 in UV 11.3.1.  However, on 14.2.1, it acts like a simple “R#8” and does not add the leading zeroes.Does anyone have an idea why this functionality changed and what to do to correct it?  This could damage many of our files’ data that rely on this functionality.Nelson Schroth

Greetings, Rocketeers!How can I accomplish moving BASIC source code from a production environment to a development environment?…THX!~Doc

I'm trying to get ssh working into a Windows 2022 system with UniVerse via openssh. It works much better than the last time I tried it, but for one problem: the escape key. The problem is that the escape key is not passed to UniVerse. I'm wondering if anyone has any suggestions or has managed to get this going? Let me illustrate this a little. If I ssh into the Windows server, CD to a UniVerse account, and run the uv (or uvsh) command, and then run the following program: GETKEY 0001 PRINT 'This will list the ASCII code of the key pressed until you hit "Q"' 0002 loop 0003 c = keyin() 0004 print seq(c) 0005 until c = 'Q' ! c = 'q' do repeat It returns the ASCII code for any key I press - except for the ESC key, when it simply doesn't respond. Curiously, the same problem occurs if I run the program after opening a UniVerse command line session using Start > Rocket U2 > UniVerse Shell (so the problem isn't exclusively related to ssh, it would seem). Furthermore, if from th

The code below suddenly stopped working. I perform the same request with a curl statement and it works.  does someone see something I should change or add to my request or is universe 11.xx not up to date with modern security? 051:       HTTP.METHOD    = "POST"052:       OUT.URL        = TOKEN.URL:TOKEN.ENDPOINT053:       VERSION.CODE   = setHTTPDefault('VERSION','1.1')054:       CTX.TYPE       = ''055:       MY.RTN.CODE    = createSecurityContext(MY.CONTEXT, CTX.TYPE)056:       CREATE.CODE    = createSecureRequest(OUT.URL,HTTP.METHOD,HANDLE,MY.CONTEXT)057:       RET1 = setRequestHeader(HANDLE,'Accept','*/*')058:       RET1 = setRequestHeader(HANDLE,"content-type","application/x-www-form-urlencoded")059:       RET1 = setRequestHeader(HANDLE,'HTTP Protocol','HTTP/1.1')060:       REQUEST.OUT  = 'grant_type=password'061:       REQUEST.OUT:= '&username=':USERNAME062:       REQUEST.OUT:= "&password=":USERPW063: *064:       stat           = submitRequest(HANDLE,30000,REQUEST.OUT,RESP

Hi everyone, been a while since I visited and even longer since I shared.We are moving from SQL file triggers to the IDX file triggers and ran into a small issue where we have a trigger on the DICT and a trigger on the DATA levels of the same file.Getting the DATA level trigger working is simple but how to get the trigger in the DICT working?Not sure if anyone has done this before but in short the answer is you need two dictionary items: INDEX.ITYPE for the data level and D_INDEX.ITYPE for the dictionary level. To build the index on D_INDEX.ITYPE you need to use the “USING” clause in the CREATE.INDEX command:CREATE.INDEX DICT FILENAME D_INDEX.ITYPE USING DICT FILENAME NO.NULLSThe trigger will now fire when a record is added/deleted/modified in DICT FILENAME.

Universe 11.3.1,  AIX 7200-05-11-2546We noticed when comparing very large numbers they don’t correctly evaluate to not equal. Specifically, if two variables are over 17 characters in length and they don’t match correctly. See the below sample program and its output. The (X NE Y) does not evaluate to true until the length of the strings are 17 characters even though the variables are never equal. Does anyone know what is causing this to happen?  X = '0587634906513946593645'Y = '0587634906513946593646' FOR Z = 1 TO 8  CRT X, Y, (X NE Y),LEN(X)  STRLN = LEN(X)  X = RIGHT(X,STRLN-1)  Y = RIGHT(Y,STRLN-1)NEXT ZOutput:0587634906513946593645        0587634906513946593646        0         22        587634906513946593645         587634906513946593646         0         21        87634906513946593645          87634906513946593646          0         20        7634906513946593645 7634906513946593646 0         19                            634906513946593645  634906513946593646  0         18        

A recent client asked how we produce different type documents within in the Universe/Unidata database. There is no Rocket Software product that does that I am aware of. Using a Eclipse open-source product called BIRT with a Universe/Unidata ODA interface you can produce all of those types and more. So, am I right about no Rocket Software product that creates reports in different formats?

I can’t remember the initiation command to use instead of “uv” to enable multiple concurrent sessions.  Can someone help me?Thanks, Dale

Hi everyone, Does anyone know how to make the "Universe Telnet Service" that runs on our IBS Application server, to only use 128bitblock Ciphers, and disable the other lower 64bit block ciphers? The service listens on port 992. The problem is that our Qualys VUN scanning tool is picking up that port 992 is a security risk againstBirthday attacks against TLS ciphers with 64bit block size vulnerability (Sweet32) and goes on to say"Disable and stop using DES, 3DES, IDEA or RC2 ciphers.  We believe that the "Universe Telnet Service" uses OpenSSL that offers out these lower level ciphers, but we have no idea how to to best address this sothat the only ciphers used are 128-bit or higher ciphers. We are waiting on our third party support to get back to us as they have opened a ticket today. Current version of OpenSSL is OpenSSL 1.0.2m-fips  2 Nov 2017, though "Programs and Features" is saying OpenSSL 1.1.1k (64-bit) was installed on March 2021. Universe is 12.1.1 is installed. Thanks in advan

Michael Rajkowski March 29, 2017 Rocket Software has added support for the Python development language to Rocket® UniVerse 11.3.1 and Rocket® UniData 8.2, on Linux and Windows only. Python will also be added to Rocket® D3 in the second half of 2017. Adding Python support gives experienced and new MultiValue developers the opportunity to code in one of the most popular development languages available today. To help developers get started Rocket created a Python community page that hosts links to code samples, documentation and lots more. PYTHON IN MULTIVALUE – THE BASICS Python now sits at the same level as the other access method used by the MultiValue developer. Our MultiValue Python module consists of two parts. The main access point for the Python developer is the u2py module, which is Python code. The u2py module calls a module written in C. Since both Python and the U2 databases are written in C, it is compatible and can be embedded into the database. The entire box represents