Cluster replica

Create and manage cluster replicas.

Part of the API reference collection

This page is part of the Crunchy Bridge API reference, and primarily meant to act as an exhaustive guide for technical integrations which are already in progress. To understand the basics of using the API, see API concepts and getting started.

The `Cluster` API resource

A database cluster.

Content type: application/json

Name	Nullable	Type	Description
`id`		`string` in EID format in UUID format	The unique ID of the cluster.
`cluster_id`	✔	`string` in EID format in UUID format	In the case of a replica, the ID of the primary cluster. Always `null` for primaries and forks. Deprecate: Prefer use of `parent_id`.
`cpu`		`integer`	The number of CPU units on the cluster's instance.
`created_at`		`string` of date/time formatted as RFC 3339 in EID format in UUID format	Time at which the cluster was originally created.
`disk_usage`		`object` of type `ClusterDiskUsageobject`	Contains information on the cluster's current disk usage. This data is collected about once a day, and isn't guaranteed to be perfectly fresh. For a new cluster, it may also not have been collected yet at all, in which case this field will be `null`.
`environment`	✔	enum `string`	The type of environment which the cluster belongs to. This is largely for informational purposes, but may affect the cluster's display and interaction from Bridge tools. For example, a value of `production` will show a special banner to indicate the higher risk environment. Enum `production`, or , .
`host`		`string`	The hostname of the cluster.
`is_ha`		`boolean`	Whether the cluster is high availability, meaning that it has a secondary it can fail over to quickly in case the primary becomes unavailable.
`is_protected`		`boolean`	Whether the cluster is protected. Protected clusters can't be destroyed until their protected flag is removed (which can be done via the "update cluster" endpoint).
`is_suspended`		`boolean`	Indicates whether the cluster is currently in a suspended state, meaning that while its disk is available, it doesn't currently have a running virtual machine.
`maintenance_window_start`	✔	`integer`	The hour of day which a maintenance window can possibly start. This should be an integer from `0` to `23` representing the hour of day which maintenance is allowed to start, with `0` representing midnight UTC. Maintenance windows are typically three hours long starting from this hour. A `null` value means that no explicit maintenance window has been set and that maintenance is allowed to occur at any time.
`major_version`		`integer`	The cluster's major Postgres version. For example, `14`.
`memory`		`number`	The total amount of memory available on the cluster's instance in GB (gigabytes). This is normally a whole integer, but may be fractional for the case of smaller hobby instances that have less than a gigabyte of memory.
`name`		`string`	A human-readable name for the cluster.
`network_id`	✔	`string` in EID format in UUID format	The ID of the network where the cluster is located.
`parent_id`	✔	`string` in EID format in UUID format	The ID of the primary or parent cluster or a replica cluster. Always `null` for primaries and forks.
`plan_id`		`string`	The ID of the cluster's plan. Determines instance, CPU, and memory.
`postgres_version_id`		`string` in EID format in UUID format	The ID of the cluster's major Postgres version.
`provider_id`		`string`	The cloud provider where the cluster is located.
`region_id`		`string`	The provider region where the cluster is located.
`replicas`		`array` of `array`	Replicas for the cluster.
`storage`		`integer`	The amount of storage available to the cluster in GB (gigabytes).
`tailscale_active`		`boolean`	Indicates whether tailscale is active for the cluster.
`team_id`		`string` in EID format in UUID format	The ID of the team that the cluster belongs to.
`updated_at`		`string` of date/time formatted as RFC 3339 in EID format in UUID format	Time at which the cluster was last updated.

Example

{
    "cluster_id": null,
    "cpu": 16,
    "created_at": "2021-07-11T01:02:03Z",
    "disk_usage": {
        "disk_available_mb": 95000,
        "disk_total_size_mb": 100000,
        "disk_used_mb": 5000
    },
    "environment": null,
    "host": "p.rvf73a77ozfsvcttryebfrnlem.crunchybridge.com",
    "id": "rvf73a77ozfsvcttryebfrnlem",
    "is_ha": false,
    "is_protected": false,
    "is_suspended": false,
    "maintenance_window_start": 21,
    "major_version": 13,
    "memory": 64,
    "name": "crunchy-production",
    "network_id": "p56biajnfvgjhftvqs7lqymspe",
    "parent_id": null,
    "plan_id": "standard-64",
    "postgres_version_id": "6ibnzig5avgutdgdzei2jwvvt4",
    "provider_id": "aws",
    "region_id": "us-west-2",
    "replicas": [
        {
            "cluster_id": null,
            "cpu": 16,
            "created_at": "2021-07-11T01:02:03Z",
            "disk_usage": null,
            "environment": null,
            "host": "",
            "id": "drsve2t4f5e5znkmef6hmchy2e",
            "is_ha": false,
            "is_protected": false,
            "is_suspended": false,
            "maintenance_window_start": null,
            "major_version": 13,
            "memory": 64,
            "name": "crunchy-production-replica",
            "network_id": "p56biajnfvgjhftvqs7lqymspe",
            "parent_id": null,
            "plan_id": "standard-64",
            "postgres_version_id": "6ibnzig5avgutdgdzei2jwvvt4",
            "provider_id": "aws",
            "region_id": "us-west-2",
            "replicas": null,
            "storage": 100,
            "tailscale_active": false,
            "team_id": "eaevtjiudzeq7bsqbbpiscund4",
            "updated_at": "2021-07-11T01:02:03Z"
        }
    ],
    "storage": 100,
    "tailscale_active": false,
    "team_id": "eaevtjiudzeq7bsqbbpiscund4",
    "updated_at": "2021-07-11T01:02:03Z"
}

The `ClusterDiskUsage` object

Contains information about a cluster's disk usage.

Name	Type	Description
`disk_available_mb`	`integer`	Available disk left in MB (megabytes).
`disk_total_size_mb`	`integer`	Total disk size in MB (megabytes).
`disk_used_mb`	`integer`	Amount of disk currently in use in MB (megabytes).

Create cluster replica

Create a read-replica for an existing database cluster.

POST /clusters/{cluster_id}/replicas

Request

Path parameters

cluster_id: The ID of the cluster to create a replica for.

Request body schema

Content type: application/json

Name	Type	Description
`name`	`string`	A human-readable name for the replica.
`network_id`	`string` in EID format in UUID format	The ID of the network where the new replica should be created. If left empty, then the cluster is created in the same network as its primary or a new network if the replica is to be created in a different region from the primary.
`plan_id`	`string`	The ID of the plan with which to create the replica. Determines instance, CPU, and memory. If not specified, defaults to the primary cluster's plan.
`provider_id`	`string`	The cloud provider in which to create the replica. If not specified, defaults to the primary cluster's provider. If specified then `plan_id` and `region_id` must also be specified.
`region_id`	`string`	The provider region in which to create the replica. If not specified, defaults to the primary cluster's region. If set, then either `network_id` must not be set or `network_id` must be a network that exists in the specified region.

Example request body

{
    "name": "crunchy-production-replica",
    "plan_id": "standard-64"
}

cURL example

curl -X post https://api.crunchybridge.com/clusters/{cluster_id}/replicas
    -H "Authorization: Bearer $CRUNCHY_API_KEY"
    -H "Content-Type: application/json"
    -d '{"name":"crunchy-production-replica","plan_id":"standard-64"}'

Response

Status: 201

Responds with the standard Cluster API resource.

Destroy cluster replica

Delete a read-replica database cluster.

Deprecated: It's recommended to use the normal cluster destroy endpoint instead of this one. All clusters are uniquely addressable with their ID, so having the ID of their parent cluster in the URL as well conveys no benefit, and the semantics of this endpoint lose cohesion when looking at a replica of a replica, which is allowed. If the immediate parent of the replica of a replica is in the URL, then it stands to reason that its parent should be in the URL as well, but it's not. Using the normal cluster destroy endpoint with a single unique cluster ID produces the same result minus the semantic issues.

DELETE /clusters/{cluster_id}/replicas/{replica_id}

Request

Path parameters

cluster_id: Unique ID of the cluster which the replica belongs to.
replica_id: Unique ID of the replica to destroy.

cURL example

curl -X delete https://api.crunchybridge.com/clusters/{cluster_id}/replicas/{replica_id}
    -H "Authorization: Bearer $CRUNCHY_API_KEY"

Response

Status: 200

Responds with the standard Cluster API resource.

Detach replica

Detach a replica from its primary.

The replica detaches asynchronously and calling this endpoint multiple times before the final detach resolves will no-op idempotently. Calling detach on a non-replica is an error.

PUT /clusters/{cluster_id}/actions/detach

Request

Path parameters

cluster_id: The ID of the cluster to detach.

cURL example

curl -X put https://api.crunchybridge.com/clusters/{cluster_id}/actions/detach
    -H "Authorization: Bearer $CRUNCHY_API_KEY"

Response

Status: 200

Responds with the standard Cluster API resource.