We have launched the new Catalyst Cloud Metrics Service, in open beta.
The Metrics Service increases the observability of your resources running in the cloud by exposing a number of useful metrics. With its comprehensive query API you can monitor your cloud resources, and get time-series data suitable for making dashboards.
Along with metrics, a history of the state of resources is also kept for up to 90 days (with decreasing granularity), allowing a look into historical data in addition to monitoring the current state of resources.
The Metrics Service is free to use for all Catalyst Cloud customers. To find out more, see Catalyst Cloud Metrics Service.
The Catalyst Cloud Alarm Service now uses the Metrics Service as
the source for resource metrics. Threshold alarms monitoring resource
metrics can be created using the gnocchi_resources_threshold and
gnocchi_aggregation_by_resources_threshold alarm types.
For more information, see the documentation for the Catalyst Cloud Alarm Service.
When configuring auto-scaling with the Catalyst Cloud Orchestration Service, alarms are now configured using one of the types backed by the Metrics Service listed above.
For more information, see the documentation for auto-scaling with the Catalyst Cloud Orchestration Service.
Note
These changes do not apply to Catalyst Cloud Kubernetes Service (CCKS) clusters running Kubernetes v1.28 and later, as newer clusters use a different mechanism for auto-scaling.
If you have any CCKS clusters running Kubernetes v1.27 or earlier, please consider rebuilding your cluster to upgrade to the latest version of Kubernetes and ensure your workloads continue working correctly.