Migration services

Pentaho and Kettle to Apache Hop migration

Your .ktr and .kjb files are closer to Hop than they look. We inventory the repository, import it with Hop's tooling, fix what the import could not map, prove the results match, and cut over on your schedule.

Migration

Moving from Pentaho and Kettle to Apache Hop

Apache Hop began in 2019 as a fork of Kettle, the engine inside Pentaho Data Integration. Your transformations and jobs are closer to Hop than they look, and most of the move is mechanical. The work is in the edges: unsupported plugins, scripting steps, schedules, and proving the results match.

Why teams move

  • Pentaho's community edition no longer receives releases, so self-hosted teams face a commercial license or a migration
  • Hop is governed by the Apache Software Foundation, released regularly, and built for containers rather than retrofitted
  • The move is the natural moment to adopt Git, environments, secrets management, and monitoring instead of carrying old habits across

What your Kettle concepts become

In Kettle / PentahoIn Apache Hop
Transformation (.ktr)Pipeline (.hpl)
Job (.kjb)Workflow (.hwf)
StepTransform
Job entryAction
Spoon / PDI clientHop GUI, or Hop Web in a browser
Pan and Kitchenhop-run
Carte serverHop Server, in a container
File or database repositoryA project in Git
kettle.properties, shared.xmlProject metadata and environments
Named parameters and variablesVariables, resolved per environment

How we migrate

  1. 1

    Inventory

    Scan the repository: every transformation and job, the steps and plugins they use, connections, variables, schedules, and what still runs. This is where the surprises get found, not at cutover.

  2. 2

    Automated import

    Hop's import tool converts .ktr and .kjb files into pipelines and workflows and turns connections and variables into project metadata. It also reports every step it could not map.

  3. 3

    Remediation

    Replace unsupported or Pentaho-specific steps, rework scripting steps, and fix the patterns that only worked because of a repository. AI assistants make the repetitive parts fast; engineers own the judgment calls.

  4. 4

    Verification

    Run the old and new pipelines side by side against the same inputs and compare outputs row by row: counts, checksums, and spot checks, with the comparison scripted so it can be repeated.

  5. 5

    Cutover

    Move schedules onto Hop workflows, retire Carte in favor of containerized Hop Server, switch monitoring over, and keep the Pentaho environment idle but intact until you are satisfied.

KETTLE / PDI.ktr, .kjb, shared.xml,kettle.properties,repository, schedulesHOP IMPORTconverts files,connections, variablesreports what it cannot mapHOP PROJECT IN GIT.hpl, .hwf, metadata,environments fordev, test, productionSIDE BY SIDEsame inputs, both enginesrow counts, checksums,scripted and repeatableCUTOVERinventoryimportrun bothmatchREMEDIATIONunsupported steps, scripting, repository-only patterns,AI-assisted for the repetitive parts, engineer-reviewedgap reportfixed files committedmismatch: back to remediationPentaho stays intact and idle until the comparison is clean and the schedules have moved.
The migration path: the Kettle repository is inventoried and imported into a Hop project in Git, the import's gap report drives remediation, and a scripted side-by-side comparison against the old outputs gates the cutover.

Questions

Migration questions we hear most

Is Apache Hop the same as Kettle or Pentaho Data Integration?
No, but they are close relatives. Apache Hop began in 2019 as a fork of Kettle, the engine inside Pentaho Data Integration, and is now a separate project governed by the Apache Software Foundation. The concepts map directly: transformations become pipelines, jobs become workflows, steps become transforms, and job entries become actions.
Can existing .ktr and .kjb files be imported into Hop?
Yes. Hop's import tool converts .ktr and .kjb files into pipelines and workflows and turns connections and variables into project metadata. It also reports every step it could not map, which is where the remediation work begins: unsupported or Pentaho-specific steps, scripting steps, and patterns that only worked because of a repository.
Why move off Pentaho Data Integration?
Pentaho's community edition no longer receives releases, so self-hosted teams face a commercial license or a migration. Hop is released regularly under the Apache Software Foundation and was built for containers. The move is also the natural moment to adopt Git, environments, secrets management, and monitoring.
How do we know the migrated pipelines produce the same results?
We run the old and new pipelines side by side against the same inputs and compare the outputs row by row: counts, checksums, and spot checks. The comparison is scripted so it can be repeated, and it gates the cutover.
Do we have to migrate everything at once?
No. Pipelines move in slices, schedules move onto Hop workflows as each slice is verified, and the Pentaho environment stays idle but intact until you are satisfied. There is no big-bang rewrite.

Next step

Free Consultation

Tell us about your data platform, or the one you wish you had. A short call is usually enough to say whether we can help and what it would take.

Get in touch(877) 881-6238