Greenplum decision guide
Greenplum: stay, upgrade or exit.
Broadcom's 2024 decision changed what you can assume about Greenplum. It did not force anyone off it. This guide helps you decide with your own constraints.
The short answer
Broadcom closed the Greenplum open-source repository in 2024 and now ships it as a commercial product. You have three paths: stay on supported Greenplum, upgrade within it, or exit to Apache Cloudberry, another Postgres-based MPP, or a cloud warehouse such as Snowflake or Databricks. The right path depends on licensing, workload and skills.
Background
What changed with Greenplum in 2024?
Greenplum was an open-source massively parallel processing database built on PostgreSQL. In 2024 Broadcom, which owns it through VMware, closed the public source repository and moved Greenplum to a commercial-only model under the Tanzu name. Teams that relied on community builds or on the code being open lost that option.
The community response was a fork. Apache Cloudberry, a project at the Apache Software Foundation, continues from the Greenplum codebase as open source. Other vendors have also built Greenplum-derived products. See Greenplum alternatives for the options.
Greenplum itself still runs, and a support path exists. Nothing about the closure stops your cluster tonight. What changes is your long-term position on licensing, roadmap and community.
Options
What are the three paths?
| Path | What it means | Best when | Main cost |
|---|---|---|---|
| Stay | Keep Greenplum under a support agreement | Workload is stable, the licence is acceptable, nobody wants a migration | Recurring licence and a closed roadmap you do not control |
| Upgrade | Move to a current Greenplum release | You run Greenplum 5 or 6 and need a supported version | Upgrade effort, plus all the stay costs |
| Exit | Move to Cloudberry, another Postgres-based MPP, or a cloud warehouse | Licence cost, cloud strategy or open-source requirements drive the decision | Migration effort and re-testing |
Path 1 and 2
When does staying on Greenplum make sense?
Stay if the workload is stable, the cluster meets its service levels, and the licence cost is smaller than the cost and risk of a migration. A migration is a project with its own failure modes. For a warehouse that is not growing, a support contract is a rational choice.
Upgrade first if you are on an older major version. Running Greenplum 5 or 6 is a separate risk from the licence question, because older versions leave support windows. AceMQ licenses and supports Greenplum, and can price the stay path against an exit. That is why this page links to AceMQ's Greenplum pages for the stay case: Ace8 owns the exit decision, AceMQ owns Greenplum support.
Path 3
When should you exit Greenplum?
Where to go next depends on how much of your SQL, external tables and ETL is Greenplum-specific. The closer the target is to Greenplum, the smaller the rewrite.
- The licence cost does not justify the value, and you have a number to prove it.
- Open source is a requirement, because of policy, contract or portability concerns.
- The business is moving data to a cloud warehouse or lakehouse and the cluster is a leftover.
- The cluster hardware is aging, so a refresh is coming either way.
- You cannot hire engineers who want to run Greenplum.
Method
How do you decide between stay and exit?
- 1
Price the stay path
Get a current quote for supported Greenplum at your core count and version, plus the upgrade if needed.
- 2
Inventory what is Greenplum-specific
Distribution keys, append-optimized tables, external tables, gpload jobs, resource groups, PL code and user-defined functions.
- 3
Price each exit
Estimate conversion and re-testing for Cloudberry, WarehousePG and a cloud warehouse.
- 4
Compare over three to five years
Include migration labour, dual running and the new platform's run cost.
- 5
Decide per workload
Some workloads may stay while others move.
Risk
What are the risks of doing nothing?
Doing nothing is a decision. If you run an unsupported build, security fixes stop flowing to you. If you run supported Greenplum, your renewal is the leverage point: with no credible alternative, there is nothing to negotiate with. Even teams who stay benefit from knowing what an exit would cost. See the modernization assessment for how that is estimated.
Next step
Where to go from here
If you are leaning toward exit, read Greenplum alternatives and Greenplum to Cloudberry migration. For the broader set of database exits, see database modernization.
After go-live
Keep it running after cutover.
Broadcom closed the public Greenplum repository in 2024. Apache Cloudberry is the open-source fork that continues from the Greenplum codebase.
No. Greenplum is still sold and supported commercially. Older major versions reach their own end of support, so check the version you run.
An open-source MPP database at the Apache Software Foundation, forked from Greenplum and built on PostgreSQL. It is the closest fit for Greenplum workloads.
No. Many teams will stay under a support agreement. Leave only when your numbers or requirements say so.
Apache Cloudberry, and Greenplum-derived products such as EDB WarehousePG. Cloud warehouses like Snowflake and Databricks are further from Greenplum and need more rewriting.
A move to Cloudberry usually needs less code change than a move to a cloud warehouse. Time depends on object count, external tables and ETL jobs. The assessment gives a range.
Know the cost before anything moves.
A fixed-price assessment gives you the inventory, the risk score and a phased plan. Scope and cost are fixed before the migration starts.
Book a Fixed-Price AssessmentGet in touch
Talk to a modernization engineer
Tell us what you run today. An engineer, not a sales script, replies within one business day.