Managing Clusters

Clusters are first-class citizens in Senlin service design. A cluster is defined as a collection of homogeneous objects. The “homogeneous” here means that the objects managed (aka. Nodes) have to be instantiated from the same “profile type”.

List Clusters

To examine the list of receivers:

def list_cluster(conn):
    print("List clusters:")

    for cluster in conn.clustering.clusters():
        print(cluster.to_dict())

    for cluster in conn.clustering.clusters(sort='name:asc'):
        print(cluster.to_dict())

When listing clusters, you can specify the sorting option using the sort parameter and you can do pagination using the limit and marker parameters.

Full example: manage cluster

Create Cluster

When creating a cluster, you will provide a dictionary with keys and values according to the cluster type referenced.

def create_cluster(conn):
    print("Create cluster:")

    spec = {
        "name": CLUSTER_NAME,
        "profile_id": PROFILE_ID,
        "min_size": 0,
        "max_size": -1,
        "desired_capacity": 1,
    }

    cluster = conn.clustering.create_cluster(**spec)
    print(cluster.to_dict())

Optionally, you can specify a metadata keyword argument that contains some key-value pairs to be associated with the cluster.

Full example: manage cluster

Get Cluster

To get a cluster based on its name or ID:

def get_cluster(conn):
    print("Get cluster:")

    cluster = conn.clustering.get_cluster(CLUSTER_ID)
    print(cluster.to_dict())

Full example: manage cluster

Find Cluster

To find a cluster based on its name or ID:

def find_cluster(conn):
    print("Find cluster:")

    cluster = conn.clustering.find_cluster(CLUSTER_ID)
    print(cluster.to_dict())

Full example: manage cluster

Update Cluster

After a cluster is created, most of its properties are immutable. Still, you can update a cluster’s name and/or params.

def update_cluster(conn):
    print("Update cluster:")

    spec = {
        "name": "Test_Cluster001",
        "profile_id": "c0e3a680-e270-4eb8-9361-e5c9503fba0a",
        "profile_only": True,
    }
    cluster = conn.clustering.update_cluster(CLUSTER_ID, **spec)
    print(cluster.to_dict())

Full example: manage cluster

Delete Cluster

A cluster can be deleted after creation, When there are nodes in the cluster, the Senlin engine will launch a process to delete all nodes from the cluster and destroy them before deleting the cluster object itself.

def delete_cluster(conn):
    print("Delete cluster:")

    conn.clustering.delete_cluster(CLUSTER_ID)
    print("Cluster deleted.")

    # cluster support force delete
    conn.clustering.delete_cluster(CLUSTER_ID, False, True)
    print("Cluster deleted")

Add Nodes to Cluster

Add some existing nodes into the specified cluster.

def add_nodes_to_cluster(conn):
    print("Add nodes to cluster:")

    node_ids = [NODE_ID]
    res = conn.clustering.add_nodes_to_cluster(CLUSTER_ID, node_ids)
    print(res)

Remove Nodes from Cluster

Remove nodes from specified cluster.

def remove_nodes_from_cluster(conn):
    print("Remove nodes from a cluster:")

    node_ids = [NODE_ID]
    res = conn.clustering.remove_nodes_from_cluster(CLUSTER_ID, node_ids)
    print(res)

Replace Nodes in Cluster

Replace some existing nodes in the specified cluster.

def replace_nodes_in_cluster(conn):
    print("Replace the nodes in a cluster with specified nodes:")

    old_node = NODE_ID
    new_node = "cd803d4a-015d-4223-b15f-db29bad3146c"
    spec = {old_node: new_node}
    res = conn.clustering.replace_nodes_in_cluster(CLUSTER_ID, **spec)
    print(res)

Cluster Scale Out

Inflate the size of a cluster.

def scale_out_cluster(conn):
    print("Inflate the size of a cluster:")

    res = conn.clustering.scale_out_cluster(CLUSTER_ID, 1)
    print(res)

Cluster Scale In

Shrink the size of a cluster.

def scale_out_cluster(conn):
    print("Inflate the size of a cluster:")

    res = conn.clustering.scale_out_cluster(CLUSTER_ID, 1)
    print(res)

Cluster Resize

Resize of cluster.

def resize_cluster(conn):
    print("Resize of cluster:")

    spec = {
        'min_size': 1,
        'max_size': 6,
        'adjustment_type': 'EXACT_CAPACITY',
        'number': 2,
    }
    res = conn.clustering.resize_cluster(CLUSTER_ID, **spec)
    print(res)

Attach Policy to Cluster

Once a policy is attached (bound) to a cluster, it will be enforced when related actions are performed on that cluster, unless the policy is (temporarily) disabled on the cluster

def attach_policy_to_cluster(conn):
    print("Attach policy to a cluster:")

    spec = {'enabled': True}
    res = conn.clustering.attach_policy_to_cluster(
        CLUSTER_ID, POLICY_ID, **spec
    )
    print(res)

Detach Policy from Cluster

Once a policy is attached to a cluster, it can be detached from the cluster at user’s request.

def detach_policy_from_cluster(conn):
    print("Detach a policy from a cluster:")

    res = conn.clustering.detach_policy_from_cluster(CLUSTER_ID, POLICY_ID)
    print(res)

Cluster Check

Check cluster health status, Cluster members can be check.

def check_cluster(conn):
    print("Check cluster:")

    res = conn.clustering.check_cluster(CLUSTER_ID)
    print(res)

Cluster Recover

To restore a specified cluster, members in the cluster will be checked.

def recover_cluster(conn):
    print("Recover cluster:")

    spec = {'check': True}
    res = conn.clustering.recover_cluster(CLUSTER_ID, **spec)
    print(res)