@kapa.ai We have deployed Airbyte to our EKS cluster using helm chart. What are the minimum resources (RAM+CPU) requirements for connect builder server, workload api server?

The knowledge sources do not define official minimum CPU/RAM requirements for the connector-builder-server or workload-api-server. They only show example configurations and note that resource sizing is left to the operator. From the docs for the Helm

values.yaml

, Airbyte explicitly recommends not setting defaults and instead letting users choose resources based on their environment. [[Deploying values](https://docs.airbyte.com/platform/deploying-airbyte/values)] For both

connector-builder-server

and

workload-api-server

, the default chart values have:

yaml resources:   limits: {}   requests: {}

meaning no hard minimum is enforced. [[Deploying values](https://docs.airbyte.com/platform/deploying-airbyte/values)] However, there are some example configurations from issues and sample values that can serve as rough starting points (not official minimums): - Connector Builder Server - Example 1 (GitHub issue):

yaml     connector-builder-server:       resources:         requests:           memory: 256Mi           cpu: 200m         limits:           memory: 1Gi           cpu: 500m

[[Worker bean issue](https://github.com/airbytehq/airbyte/issues/60915)] - Example 2 (Temporal on EKS issue):

yaml     connector-builder-server:       resources:         requests:           memory: 350Mi           cpu: 10m         limits:           memory: 500Mi           cpu: 200m

[[Temporal deadline EKS](https://github.com/airbytehq/airbyte/issues/59730)] - Workload API Server - Chart defaults: no requests/limits set (left empty). [[Deploying values](https://docs.airbyte.com/platform/deploying-airbyte/values)] - Example (Temporal on EKS issue):

yaml     workload-api-server:       resources:         requests:           memory: 300Mi           cpu: 10m         limits:           memory: 800Mi           cpu: 500m

[[Temporal deadline EKS](https://github.com/airbytehq/airbyte/issues/59730)] Given the lack of official minimums in the docs, you can: 1. Start with something in the range of the examples above (e.g. ~256–350Mi RAM request and 200m CPU for connector-builder-server; ~300Mi RAM and small CPU request for workload-api-server). 2. Monitor pod usage and adjust requests/limits based on actual load and cluster capacity. If you need strict guidance (e.g. for very small EKS nodes), the current knowledge sources do not provide more precise minimums.