Skip to content

v5.7.8

This release addressed a query-time race condition primarily impacting Grafana alert users and addressed bugs causing queries to ignore the max_execution_time circuit breaker. This release also lowered a default socket timeout in ClickHouse.

This release contains bug fixes to v5.7.4. Refer to the release notes for v5.7.4 to see other notable feature announcements and information.

prod scale profile users may need to scale turbine-api

By default in versions v5.7.x, the prod scale profile has the following settings:

  • The turbine-api pods have 1Gi of memory and 1 CPU.
  • The turbine_api_worker_count tunable is set to 8 workers. This value may be overridden in the Hydrolix spec configuration.

For clusters using the prod scale profile, make the following adjustments to the scale of the turbine-api pod in the Hydrolix spec configuration based on the value of the turbine_api_worker_count tunable:

turbine-api scaling guidance:

  • 8 workers: A minimum of 1.7Gi memory and 1 CPU
  • 16 workers: A minimum of 2.74Gi memory and 2 CPU
  • 32 workers: A minimum of 6.80Gi memory and 3 CPU

Upgrade⚓︎

Do not skip versions when upgrading or downgrading

Skipping versions when upgrading or downgrading Hydrolix can result in database schema inconsistencies and cluster instability. Always upgrade or downgrade sequentially through each version.

Upgrade on GKE⚓︎

1
kubectl apply -f "https://www.hydrolix.io/operator/v5.7.8/operator-resources?namespace=${HDX_KUBERNETES_NAMESPACE}&gcp-storage-sa=${GCP_STORAGE_SA}"

Upgrade on EKS⚓︎

1
kubectl apply -f "https://www.hydrolix.io/operator/v5.7.8/operator-resources?namespace=${HDX_KUBERNETES_NAMESPACE}&aws-storage-role=${AWS_STORAGE_ROLE}"

Upgrade on LKE⚓︎

1
kubectl apply -f "https://www.hydrolix.io/operator/v5.7.8/operator-resources?namespace=$HDX_KUBERNETES_NAMESPACE"

Changelog⚓︎

Improvements⚓︎

Core improvements⚓︎

  • Lowered the socket timeout setting in ClickHouse from 1000 to 20 seconds to improve cancel response time on queries.

Bug Fixes⚓︎

Core fixes⚓︎

  • Fixed a bug causing the circuit breaker max_execution_time to not time out queries in the case of the queries requiring blocking operations.

Operator fixes⚓︎

  • Fixed a race condition which caused intermittent invalid token errors from ClickHouse. This bug particularly affected users of Grafana alert queries.