Product Documentationsys.cluster
This system view returns state information about each compute cluster in the current instance. See compute clusters for more information on the role of compute clusters, the lifecycle of cluster nodes and the meaning of the values below.
If a compute cluster's state is SUSPENDED, the various node counts will be null.
| Column Name | Data Type | Description |
|---|---|---|
| cluster_id | uuid | UUID for the cluster. |
| cluster_name | name | Unique name of the cluster. |
| nodes | smallint | Number of nodes requested for the cluster. |
| prepared_nodes | bigint | The number of cluster nodes we have received from the cloud provider. |
| configured_nodes | bigint | The number of cluster nodes that have the database software pod and image. |
| ready_nodes | bigint | Number of cluster nodes with the database software starting up. |
| ready_workers | bigint | Number of cluster nodes that have started the software and are ready to execute queries. |
| default_wlm_profile_name | name | Name of the default WLM profile that was defined for the cluster. |
| active_wlm_profile_name | varchar(64000) | Name of the active WLM profile currently assigned to the cluster. |
| hardware_instance_type_id | uuid | UUID for the hardware instance type selected for the cluster. |
| hardware_instance_name | varchar(256) | Name of the hardware instance type selected for the cluster: small-v1 or large-v1. |
| is_default_cluster | boolean | Whether this cluster is the default cluster. See Default and System Clusters. |
| is_system_cluster | boolean | Whether this cluster is the system cluster. See Default and System Clusters. |
| auto_suspend | integer | Auto-suspend timeout, in seconds. 0 means that auto-suspend is disabled (as for the default cluster). |
| auto_resume | boolean | Whether this cluster resumes automatically when a query or some other operation attempts to use it. |
| max_spill_pct | integer | Maximum percentage of total disk that can be used for spill space. |
| max_cache_pct | integer | Maximum percentage of total disk that can be used for the object store cache. |
| type | text | Always COMPUTE (virtual compute cluster). |
| state | text | Cluster state: CREATING, RUNNING, SUSPENDED. |
Examples
Show the number of nodes we've requested but not yet received from the cloud provider. This can show lack of free nodes in a certain availability zone, or account quota issues:
SELECT cluster_name, nodes - prepared_nodes from sys.cluster;