This was implemented by the maintainers: the Helm chart now exposes the concurrency block independently of the queued run coordinator, so you can set concurrency pools through additionalInstanceConfig without a custom run coordinator. If you are on an older chart version and hit the DagsterInvalidConfigError about max_concurrent_runs, that is the conflict: either upgrade the chart, or work around it by setting the concurrency config only in the instance YAML while leaving the run coordinator on defaults.

Context: GitHub issue dagster-io/dagster#31086 (closed, 9 comments): setting the concurrency pools block via additionalInstanceConfig in the Dagster Helm chart failed daemon startup with DagsterInvalidConfigError, because the Helm run-coordinator template always injected max_concurrent_runs, which is incompatible with the concurrency > runs config. There was no way to leave that field unset through Helm values, so concurrency limits could not be configured without defining a fully custom run coordinator.

## Matched source
Source: Published skill
Original query: "Dagster Helm concurrency pools blocked by injected max_concurrent_runs setting"
Key terms: blocked, concurrency, concurrent, dagster, helm, injected, pools, runs, setting
