Your OpenVMS migration, run by AI. Verified to the last record.

Our look-alike tools make the conversion deterministic: we refine the tools, never your application code. On top of them, AI specially trained on OpenVMS does the slow part, fast.

Talk to an engineer

Stay on OpenVMS, move to x86

The Rdb toolchain Oracle left behind

Oracle declined to port Rdb and CDD to x86 OpenVMS. Keep your application and your build as they are: these drop-in replacements answer to the names you already use, VX/CDD supplies the dictionary, and the database underneath becomes PostgreSQL, SQL Server, MySQL or Oracle.

SQL$PRE (VX/SQL)

Drop-in for the Rdb embedded-SQL precompiler: same command line, same outputs, same status codes. COBOL, FORTRAN, Pascal and C, with XA two-phase commit through DECdtm.

Proven on HP's AVERTZ ACMS application: original 1991 source, built and run on x86 against PostgreSQL.

SQL$MOD (VX/SQLMOD)

Drop-in for the Rdb SQL module language compiler.

The same tool converted all of IKEA's 600+ Rdb SQLMOD files.

RDB$PRE (VX/RDML)

Two ways to use it. As a drop-in for the Rdb RDML precompiler: your RDML builds on x86 as it does today. Or as a converter: it rewrites the RDML once into standard embedded SQL. COBOL, BASIC, FORTRAN, Pascal and C.

VX/TUXEDO (Tuxedo for x86 OpenVMS)

Wire compatible with Oracle Tuxedo 22.1.1: TDOMAIN in both directions and unmodified /Workstation clients, so a Tuxedo estate moves to x86 one domain at a time.

Faster than Oracle Tuxedo in head-to-head runs: 1.80 times the calls per second at 16 clients.

Talk to an engineer

Move to native Linux

Every layer of OpenVMS, already rebuilt for Linux

A fast, AI-driven port needs nothing invented mid-project. Each part of your system has a Sector7 tool waiting for it, so the AI drives a toolchain that already exists.

Languages

COBOL, FORTRAN, BASIC, PASCAL and C translated to standard compilers; GnuCOBOL -fdec

Rdb database

VX/SQL (COBOL, FORTRAN), VX/SQLMOD, VX/RDML: Rdb to PostgreSQL or Oracle

Transaction processing

VX/ACMS, VX/RTR, VX/DMQ

Forms and screens

VX/DECFORMS, VX/TDMS, VX/FMS, VX/SMG

Files and data

VX/RMS, VX/FDL, VX/DATAX

DCL and system

VX/DCL, print and batch queues, lock manager, SORT/MERGE, CLD

Runtime libraries

VX/RT: call-compatible LIB$, SYS$, STR$, OTS$, MTH$ and 15 more

Dictionary and build

VX/CDD, VX/MSG, VX/MMS

Start from your application

Rdb to PostgreSQL / Oracle

Talk to an engineer

Measured, not estimated

Faster than the Itanium original

ia64 running native Oracle Rdb, against OpenVMS x86 running VX/SQL to PostgreSQL. Same 8,192 input records and the same seven workloads, 57,344 queries a cycle.

ia64, native Oracle Rdb

373 s

One full cycle (median of six).

x86 OpenVMS, VX/SQL to PostgreSQL

9.9 s

The same cycle.

Per full cycle

37.8×

Faster (ratio of the two medians).

On joins, up to

105.7×

Faster: a 3-table invoice join takes 17.65 ms on ia64 Rdb, 0.17 ms on VX/SQL.

Time per query, by workload

Milliseconds per complete query, median of six runs. Shorter is faster.

Lookup, 32-byte payload 4.7× faster

0.69 ms

0.15 ms

Lookup, 256-byte payload 1.8× faster

0.26 ms

0.15 ms

Lookup, 1,024-byte payload 1.8× faster

0.27 ms

0.15 ms

Range, 64 orders 23.3× faster

4.61 ms

0.20 ms

Invoice, 3-table join 105.7× faster

17.65 ms

0.17 ms

Invoice, 5-table join 93.3× faster

17.41 ms

0.19 ms

Correlated subquery 23.9× faster

4.60 ms

0.19 ms

One full cycle of 57,344 queries

Seconds per cycle, median of six cycles. Every VX/SQL cycle took 9.561 to 10.457 s; every Rdb cycle took 372.550 to 373.583 s.

Median cycle 37.8× faster

373.0 s

9.9 s

Show the full figures

Per-query times, in milliseconds per query. t86 is OpenVMS x86 running VX/SQL to PostgreSQL on RH8X; Rdb runs on ia64.
Workload	VX/SQL, t86	Rdb, ia64	t86 speedup
Lookup, 32-byte payload	0.1465	0.6873	4.69×
Lookup, 256-byte payload	0.1472	0.2633	1.79×
Lookup, 1,024-byte payload	0.1511	0.2748	1.82×
Range, 64 orders	0.1982	4.6107	23.26×
Invoice, 3-table join	0.1669	17.6498	105.74×
Invoice, 5-table join	0.1865	17.4067	93.33×
Correlated subquery	0.1922	4.6008	23.93×

Complete cycles of 57,344 queries, in seconds.
Cycle	t86 VX/SQL	ia64 Rdb
Cycle 1	9.957	373.238
Cycle 2	9.816	372.550
Cycle 3	10.457	373.583
Cycle 4	9.915	372.926
Cycle 5	9.591	372.983
Cycle 6	9.561	372.575
Median cycle	9.866	372.955
All six cycles	59.297	2,237.855

Read-only workload. t86 is OpenVMS x86 as a KVM guest on an RH8X host, reaching PostgreSQL over virtio; hardware differs between the platforms. PostgreSQL uses prepared execution (plan_cache_mode=auto, 128 MiB shared_buffers); Rdb uses 4,096 process buffers, a different cache mechanism. Transactions including completion are included; connection setup, warm-up and reporting are excluded. Figures are the median of six block averages (block elapsed time divided by 8,192), not individual-query percentiles.

AI-orchestrated, deterministic migration

Generic AI does not know OpenVMS. Ours does, so the testing and tuning that used to tie up specialists for months takes a fraction of the time.

Talk to an engineer

Over 1500 Projects

Since 1985

PLANS THE TESTS

The AI plans the tests from your code and documentation.

BUILDS THE TEST DATA

It builds the test data and populates the files.

RUNS THE TESTS

It runs the tests on the original and the migrated system.

FIXES AND RE-TESTS

It fixes and re-tests until the outputs match exactly.

TUNES PERFORMANCE

It tunes the performance of the migrated code.

Oracle will NOT be porting Rdb to x86

They will also be cancelling support on Itanium from the end of 2027

Schedule a consultation to migrate your complex legacy software to a modern system.

Talk to an engineer

Sector7 Services

Empowering enterprises with efficient and scalable OpenVMS migration and modernization to cloud.

Migrations

Migrate OpenVMS applications to Linux, preserving functionality with minimal downtime. Sector7 ensures a smooth transition backed by unmatched expertise.
‍

Toolset Bundles

Equip your team with Sector7’s migration tools, simplifying rehosting and modernization. Our robust APIs ensure seamless integration and reliable performance across platforms.
‍