top of page
Search

Commercial Shifts Don’t Kill OSS — They Multiply It


Open source is in flux.


  • Open source projects are leaning commercial-first.

  • Bitnami pulled back community support.

  • K8s Ingress soon EOL ?

  • Redis changed its license from BSD to a dual RSAL/SSPL

  • Elastic cut down on theirs.


We’ve built a full validation process to identify the best choices for open-source technologies, always balancing three core elements: FeaturesStabilityCommunity growth


As part of this process, we’ve been running deep tests across operators, databases, gateways, and more. Thought I’d share a few highlights from our latest evaluations

Here’s the shortlist of OSS projects we trust (and why).



🗄️ Databases


MongoDB → Percona Operator

We tested multiple MongoDB OSS — the favorite of all:


  • Simplifies resource definitions via CRDs

  • Sharding CRDs included :-) yay

  • Enterprise-grade features without the lock-in

  • 🚀 Moving fast to cover even more enterprise demands

This one is a no-brainer. We’ll be watching more Percona database operators closely.


was simple automating Sec, shard & db relationships with Percona CRDs...


ree

PostgreSQL → CloudNativePG (CNPG)


Tons of good options for postgres, but CNPG felt super strong.

  • CRDs that make DB management smooth

  • Backup + versioning baked in

  • Feels like running Postgres on autopilot


We haven’t tested Percona Postgres yet, but CNPG is already a pro-grade solution.


Restore Cluster , Versining OOTB and more.. CNPG kick as.....

ree

📦 Object Storage → MinIO


This was a big decision. Official operator vs. community operator.The community Helm chart literally says:

“We don’t guarantee support. Use the official MinIO operator for production.”

not scary though. we added some extras

  • Jobs to manage/duplicate buckets (CRD for buckets doesn’t exist yet)

  • Integrated into AI pipelines, backups, messaging sinks

👉 Works great, stable and flexible with our automation in place.




🐇 Messaging → RabbitMQ


If you moved away from Bitnami, don’t stress.The community RabbitMQ operator is excellent.

  • Single YAML install, CRDs included (HELM an option)

  • tons of capabilities

  • Stable, with strong support

  • awesome documentations!


we a lllott of automation with the Rabbitmq cluster operator...


automate queues, exchange relations , dev documentations and QA validation tests


🔑 Identity → Keycloak operator


Keycloak has quirks.

  • Limitation: you can’t delete CRDs once created

  • Our fix: CRD only for default realm, then automation for the rest with the definition json (which the operator supports)

Clean, reliable, and flexible.



🔍 Search engine → OpenSearch


When Elastic dropped Apache 2.0, we switched.

  • OpenSearch is 100% OSS under Apache 2.0

  • Stable, feature-rich, secure

  • Roadmap is 🔥 — vectors + AI search are coming fast


⚡ API Gateway (Ingress) → Apache APISIX


not sure you noticed ,but ingress is frozen. time to move to GW API.


ree
  • A killer alternative to Kong

  • Enterprise-ready, elegant, simplified route/rule management

  • With Ingress on its way out and Gateway CRD stepping in, APISIX shines as the most feature-rich, stable OSS choice

👉 If you need a serious API gateway in 2025, APISIX is it.


split traffic , rule it , weight it and filter it
split traffic , rule it , weight it and filter it


🌍 The Real problem: Compatibility


I should have mentioned this earlier — the real problem isn’t just one OSS project going commercial, it’s the fragmentation:

  • Kafka from Apache

  • Operator from Strimzi

  • Connectors from camel

  • Database from Percona


Who guarantees all of it works together? NEXT BLOG :-)


ree





 
 
 

Comments


Logo Stacktic Modernised (1).png
  • Youtube
  • LinkedIn
bottom of page