21 October 2026¶

Catalyst Cloud Metrics Service¶

Our telemetry servies have been upgraded, introducing new resource attributes and metrics to the Catalyst Cloud Metrics Service.

Instances
- New attributes
- Updated metrics
Volumes
- New attributes
Load Balancers

Instances¶

Instance resources have been updated with new resource attributes and updated metrics.

Note

Some of the new resource attributes may not be populated until a change has been made to the instance via the Compute API e.g. configuration change, restart, shelve or unshelve.

New attributes¶

The following new resource attributes have been added.

Attribute	Description
`availability_zone`	The availability zone the instance is located in.

Updated metrics¶

Some instance metrics, such as memory.available and memory.usage, now have measures recorded more accurately, including any decimal components (which were previously rounded down to the nearest whole number).

Volumes¶

Volume resources have been updated with new resource attributes.

New attributes¶

The following new resource attributes have been added.

Attribute	Description
`volume_type_id`	The ID of the volume type this volume uses.
`image_id`	The ID of the image this volume was created from, if created from an image.
`instance_id`	The ID of the instance this volume is attached to (if attached to one), or the ID of the last instance it was attached to (if no longer attached to one).

Load Balancers¶

Load balancer resources have been updated with new resource attributes and metrics.

New attributes¶

The following new resource attributes have been added.

Attribute	Description
`availability_zone`	The availability zone the load balancer is located in.
`vip_address`	The virtual IP address the load balancer is bound to.
`provider`	The provider used to provision the load balancer. On Catalyst Cloud, the `amphora` provider is used.

New metrics¶

The following new metrics have been added.

Metric	Description
`loadbalancer.provisioning`	The provisioning status of the load balancer.
`loadbalancer.operating`	The operating status of the load balancer.

Removed metrics¶

The following metrics have been removed, and replaced with new metrics.

Removed Metric	Replaced By
`network.services.octavia.loadbalancer`	`loadbalancer.provisioning`

Catalyst Cloud Kubernetes Service¶

v1.36 Cluster Template¶

Cluster Templates for Kubernetes v1.36 are now available.

The Supported CCKS Kubernetes Versions section of our documentation has more information about this release and projected supported dates of all version releases.

To get started using our CNCF Certified Kubernetes service, follow the Kubernetes Quick-Start Guide.

API Service¶

We have released an update for the API service of CCKS with new features, supporting our new v1.36 cluster template. See our list of Kubernetes versions for more information.

Added the csi_cinder_volume_binding_mode cluster label to set the VolumeBindingMode of default storage classes. This label is not set by default.
Added the server_group_policy_control_plane cluster label to control the scheduling policy of the control plane server group. Valid values are affinity, anti-affinity, soft-affinity, and soft-anti-affinity. This label is set to anti-affinity by default.
Added the server_group_policy_workers cluster label to control the default scheduling policy of server groups for worker node groups. Valid values are affinity, anti-affinity, soft-affinity, and soft-anti-affinity. This label is set to soft-anti-affinity by default.
Added the server_group_policy node group label to control the scheduling policy of the server group of the labelled node group nodes. Valid values are affinity, anti-affinity, soft-affinity, and soft-anti-affinity. This label is not set by default and follows the policy set by server_group_policy_workers if unspecified.

The following changes are available in cluster templates v1.36 and newer:

Added the certificates_rollout_before_days cluster label to support automated rollout of new control plane nodes with renewed cluster certificates at a set number of days before expiration. Values between 1 and 364 are valid or set to false to disable this feature. This label is set to 21 by default.
Added the csi_cinder_cleanup_volumes cluster label to support automatic deletion of underlying volumes of provisioned persistent volume claims when the cluster that provisioned them is deleted. Set this label to true to enable the feature. This label is set to false by default.

See Cluster Labels for a full list of available cluster labels.