Resource Types

Resources are the individual systems, services, or storage targets that Plakar Control Plane manages as part of a backup workflow. Examples of resources include S3 buckets, EC2 instances, PostgreSQL databases, virtual machines, and filesystems.

flowchart TB
  subgraph Inventories
    Managed["Managed inventory\n(Scaleway, ...)"]
    SelfManaged["Self-managed inventory"]
  end

  subgraph Resources
    R["Resource\nclass / subclass"]
  end

  subgraph Apps
    Source["Source\n(Back up content from here)"]
    Store["Store\n(Store backups here)"]
    Destination["Destination\n(Restore here)"]
  end

  Managed -->|"auto-discover"| R
  SelfManaged -->|"manually added"| R
  R --> Source & Store & Destination

All resources discovered across all inventories are available under Infrastructure > Inventories, on the Resources tab. From here a resource can be assigned to either a source or a destination app. The list can be narrowed to one inventory or filtered, as described in Filtering resources.

The actions on each resource include placing a legal hold on it. A resource under a legal hold cannot be used by any task, so backups, restores, checks, and prunes involving it are refused until the hold is removed.

Filtering resources

The resource list can be searched by name or URN and filtered by the following values. Filters combine, so a resource is listed only when it matches all of them.

• Status: the backup coverage of the resource. Protected resources are covered by a backup task. Not protected resources are not. Unmanaged resources are left out of coverage, either because they are excluded from backup coverage or because they cannot be backed up.
• Resource class: the class of the resource.
• App: the integration of a source or destination app set up on the resource. No app lists the resources with none. Store apps are not counted.
• Residency: the residency of the resource, set on it or inherited from its inventory. Not set lists the resources with none.
• Region: the region of the resource.
• Tag: one tag the resource carries.

Resource settings

Resource settings can be updated from the Resource settings tab under each resource, which is split between General, Residency and Edge pool. General holds the settings described below. Residency and Edge pool are described in Residency and Edge pools.

For managed inventories, most settings are read-only since the resource is managed by the inventory. For self-managed inventories, all settings can be modified. Backup coverage and the edge pool can be modified for any resource regardless of inventory type, since neither comes from the inventory.

Backup coverage tracks how many of your resources are protected by backups. If a resource does not need to be backed up (for example, a test database), you can exclude it from coverage using the Exclude from backup coverage. Excluded resources are omitted from protection status and coverage reporting.

Residency

Resources can only be used together when they have the same residency. A source can only be backed up to a store with the same residency, a restore can only write to a destination with the residency of its store, and a sync can only run between stores with the same residency. The residency belongs to the resource, so every app created for it takes on its residency.

A resource inherits the residency of its inventory. A residency set in the Residency section of its settings takes precedence, for a resource located in a different country from the rest of its inventory.

Edge pools

A resource assigned to an edge pool has its tasks run on an edge of that pool. A resource without a pool has its tasks run on the Control Plane.

A resource inherits the pool of its inventory. A pool set in the Edge pool section of its settings takes precedence, for a resource whose tasks have to run on different edges than the rest of its inventory. Clearing it brings back the pool of the inventory. Every app created for the resource takes on its pool.

A backup, restore or sync runs on the pool of its source or store resource, and is refused when the two resources are in different pools, since the task runs on a single edge and an edge belongs to exactly one pool. See Selecting an edge.

Resource classification

A resource is identified by its resource URN, which cannot change once the resource exists. It also has a name, a class, a subclass and tags.

The class describes the general category the resource belongs to, while the subclass identifies the specific implementation or provider. Plakar Control Plane uses this classification to determine which integrations are compatible with a resource. Configuration bundles select resources by class, subclass and tags, and policies can narrow their scope by tag. In a managed inventory, the class and tags are discovered from the provider and cannot be edited.

Endpoints and attributes

Each resource has one or more endpoints that identify where it is reached, such as a hostname, an IP address or a bucket name. In a self-managed inventory, the endpoints are entered with the resource. In a managed inventory, they are discovered from the provider, together with attributes that describe the resource, such as the S3 endpoint of a bucket or whether versioning is enabled on it. When an app is set up on the resource, the endpoints and attributes fill in the matching fields, so they do not have to be entered again.

Environment and data class tags

A resource can carry its environment and data class as tags. When a source app is created for the resource, Plakar Control Plane fills in its Environment and Data Class from those tags. Both fields can still be changed before the app is saved or even later on.

A tag is a prefix followed by the value. The prefix has a long and a short form, and either : or = separates it from the value:

• Environment: plakar:environment:<value>, plakar:environment=<value>, plakar:env:<value> or plakar:env=<value>
• Data Class: plakar:data-class:<value>, plakar:data-class=<value>, plakar:class:<value> or plakar:class=<value>

For example, a resource tagged plakar:env=production and plakar:class=database gets production as the environment and database as the data class of its source apps.

The tags can be set in two places:

• On a resource in a self-managed inventory, in its Tags field.
• On the resource at the cloud provider, as a tag or label. A managed inventory picks them up when it discovers the resource.

Supported resources

The following pages document the supported resource types and the configuration required to use each one as a source, store, or destination.