How do you migrate or upgrade IBM Maximo 7.6 to MAS 9?
The short answer
A Maximo 7.6 to MAS 9 migration has five stages: discovery of customisations, integrations and data; platform preparation on Red Hat OpenShift; data clean-up and validation; migration and regression testing in a non-production environment; and a rehearsed cutover with user training. Customisations and data quality decide the timeline more than the number of assets.
Step 1. Discovery: know what you have
Every reliable migration plan starts with an honest inventory. Before anyone sizes servers or talks dates, list:
- Java customisations, automation scripts, escalations and cron tasks
- Integrations: what connects, in which direction, through which technology (MIF, ACE, files, APIs)
- Reports (BIRT and others) and who still uses them
- Data volumes and data quality: hierarchy depth, classifications, failure codes, history
- Users and roles, which feed the AppPoints model
The output is a phased plan and a list of decisions: what to carry forward, what to rebuild with MAS configuration, and what to retire.
Step 2. Platform: prepare OpenShift and identity
MAS runs on Red Hat OpenShift. Decide early whether it will run in your data centre, in your cloud account, or as a managed service. Plan sizing, storage, backup, network and single sign-on with your IT and security teams. This is often the longest lead-time item, so start it in parallel with discovery.
Step 3. Data: clean it while you move it
A migration is the cheapest moment you will ever have to fix asset data. At a minimum, fix the asset and location hierarchy, apply classifications to critical assets, and rationalise the failure-code hierarchy. Clean problem, cause and remedy codes are what make reliability reports, and any AI, trustworthy. Our guide to cleaning failure-code data covers this step in detail.
Step 4. Migrate and test
Run the upgrade in a non-production environment first, ideally more than once:
- Rehearse the database upgrade and time it, so the cutover window is based on evidence
- Build a regression pack from real processes: create a work order, plan it, issue parts, close it with failure codes
- Test every integration end to end, including failure and retry paths
- Let key users run their own scenarios before sign-off
Step 5. Cut over and support
Agree a cutover runbook with go and no-go criteria and a rollback plan. Train users on what has changed, especially if you are introducing Maximo Mobile or new start centres. Plan for a period of heightened support after go-live, when small configuration fixes matter most.
Maximo to MAS migration checklist
| Area | Ready when |
|---|---|
| Customisations | Each one has a decision: carry, rebuild or retire |
| Integrations | Every interface is tested end to end, including errors |
| Data | Hierarchy, classifications and failure codes meet agreed quality targets |
| Platform | OpenShift sized, backed up, monitored and connected to SSO |
| Licensing | AppPoints modelled against real user roles |
| People | Users trained; support rota in place for go-live |