Hey Anton, That makes a lot of sense regarding the production readiness for K8s users today.
If it's truly additive and doesn't change defaults, I'm fine with bundling this into a 0.9.1 release. I think we should definitely do it then. On Wed, Apr 22, 2026 at 2:31 PM Anton Borisov <[email protected]> wrote: > Hi Lorenzo, > > Fair concern, but I think the framing here is narrower than LTS. This > would just be a 0.9.1 with chart-only, additive changes - no server > code changes and no change to default behavior. > > The practical value is fairly concrete: teams trying to run Fluss on > Kubernetes today often need topology spread constraints, > PodDisruptionBudgets, and external-secret wiring before their infra or > platform teams will sign off. Without these, the usual outcome of > platform review is either “fork the chart” or “come back when it’s > production-ready.” A 0.9.1 would let those teams deploy Fluss on the > clusters they already operate, rather than waiting months for 1.0. > > I also think this helps 1.0 rather than distracts from it. If more > teams are actually running Fluss on Kubernetes between now and GA, > that should give us more real production feedback feeding into the 1.0 > release, not less. > > - Anton > -- Lorenzo Affetti Senior Software Engineer @ Flink Team Ververica <http://www.ververica.com>
