[RELEASE] ScyllaDB Monitoring Stack 4.13.0 + 4.13.1

The ScyllaDB team is pleased to announce the release of ScyllaDB Monitoring Stack 4.13.0 with 4.13.1

ScyllaDB Monitoring Stack is an open-source stack for monitoring ScyllaDB based on Prometheus and Grafana. ScyllaDB Monitoring Stack 4.13.0 supports:

  • ScyllaDB 2024.x, 2025.x, and the upcoming 2026.1 release

  • ScyllaDB Manager 3.x

  • ScyllaDB Vector-Search 1.x

Related Links

Version updates for ScyllaDB Monitoring Stack 4.13.0

  • Prometheus upgraded to version 3.8.1

  • Grafana upgraded to version 12.3.1

Dashboards updates

Overview Dashboard Change

  • Disk usage was added to the top row #2705

  • Replace the requests in the top row with reads and writes #2720. The confusing requests panel was replaced by two panels, one for reads and one for writes. The reads and writes are counted at the coordinator level, which means the number of reads and writes that reached the cluster.

  • Additional information from labels was added when available: Account name and rack. #2734. An improved integration with ScyllaDB Cloud, account name, and rack are now visible in the overview dashboard.
    On-prem users can use those labels by adding them to the target files (similar to how cluster and DC are added).

  • Most of the annotations’ switches and filtering were removed, giving a cleaner look #2710, making the overview easier to understand, a minimal set of filters and annotation switches is now in use.
    This change is only applicable to the overview dashboard; the filters and annotations were not removed from the rest of the dashboards

image

Vector Search dashboard

  • Add disk, memory, and CPU graphs of the vector search nodes.

Detailed Dashboard Change

  • Add Index Cache panels to the cache section

  • Add compaction backlog panels to the compaction section #2693

  • The tablet information row is now open by default, and the heat map was removed. #2733

Operational Changes

  • Streaming progress shows both RBNO and non-RBNO #2703

  • Move to vector search version 1.x #2739

  • Switch to grafana exporter 5.x #2675

  • Native histograms GA support #2741 - Native Histograms is an advanced Prometheus feature that is now in GA, to enable it, you would need to enable it on the ScyllaDB side by setting prometheus_allow_protobuf to true.

  • Move the extra labels to a metrics #2441 - This is relevant for Cloud and DataDog Cloud integration. For performance reasons, some of the extra labels: tablets_mode, account_name, account_id, cluster_name, rack, serverExternalId, and serverId will be only on scylla_scylladb_current_version metric.