ProductPricingDocsBlog
Back to All

6 September 2023 - v3.48.10

8 months ago by David Sztykman

You can download version 3.48.10 here.

Upgrade on GKE:

kubectl apply -f "https://www.hydrolix.io/operator/v3.48.10/operator-resources?namespace=${HDX_KUBERNETES_NAMESPACE}&gcp-storage-sa=${GCP_STORAGE_SA}"

This update requires a new service account merge-controller which has the ability to connect to the cloud storage, before you run the upgrade process you need to create and add the service account by using the following:

gcloud iam service-accounts add-iam-policy-binding ${GCP_STORAGE_SA} \
--role roles/iam.workloadIdentityUser \
--member "serviceAccount:${PROJECT_ID}.svc.id.goog[${HDX_KUBERNETES_NAMESPACE}/merge-controller]" \
--project $PROJECT_ID

Upgrade on EKS:

kubectl apply -f "https://www.hydrolix.io/operator/v3.48.10/operator-resources?namespace=${HDX_KUBERNETES_NAMESPACE}&aws-storage-role=${AWS_STORAGE_ROLE}"

This update requires a new service account merge-controller which has the ability to connect to the cloud storage, before you run the upgrade process you need to create and add the service account by adding the following policy:

{
      "Effect": "Allow",
      "Principal": {
        "Federated": "arn:aws:iam::${AWS_ACCOUNT_ID}:oidc-provider/${OIDC_PROVIDER}"
      },
      "Action": "sts:AssumeRoleWithWebIdentity",
      "Condition": {
        "StringEquals": {
          "${OIDC_PROVIDER}:aud": "sts.amazonaws.com",
          "${OIDC_PROVIDER}:sub": "system:serviceaccount:${HDX_KUBERNETES_NAMESPACE}:merge-controller"
        }
      }
    }

General

  • Add HSTS support to Traefik
  • Support for Tableau authentication mechanism
  • Expose core's cloud storage http request timeouts as tunables
  • Improve speed for SIEM connector
  • Support for ARRAY JOIN for summary table
  • Change default vacuum from max age 90d to 7d
  • Adding merge-controller support but not enabled by default
  • Enable postgres backup and high availability when using crunchydata
  • Enable access logging for traefik
  • Enable prometheus metrics for merge-head
  • Disable partition auto rebalancing in redpanda
  • Core services improving logging to include pools information and more
  • Better information about raw size and uncompressed size of partition into catalog (leading to a better usage of merge)

Notable Bug Fixes

  • Support for alter works with multi-bucket
  • Fix scale_off (scale off wasn't working when using crunchydata for catalog operation)
  • Added Vaccum support for multi-bucket storage.
  • Adding support for older version of kube-state-metrics with the client-go library