cubejs-helm-charts-kubernetes
cubejs-helm-charts-kubernetes copied to clipboard
cubestore errors with 2 routers
I'm seeing the following errors on cubestore worker pods when using the default configuration of 2 routers and 2 workers with pre-aggregations defined.
2024-05-24T15:11:43.677Z ERROR [cubestore::cluster::ingestion::job_runner] pid:1 Error in processing loop: Row with id 2 is not found for JobRocksTable
These errors occur whether cubestore is configured w/ pvc or cloud storage but disappear when only one router is used. That said, the data returned from the API looks accurate. Can Cube OSS reliably be deployed with more than 1 router?