# Account Setup > Create databases, buckets and namespaces from an F# program.

From an F# program, create the storage your Worker binds to and list what the account already has. Management.Storage has the storage operations, and Management.Compute has the queue operations.

## Guestbook Database `D1CreateDatabase` returns the new database's details as a `JsonObject`. Its `uuid` field is the `database_id` of a D1 binding. ```fsharp open FSharp.CloudEdge.Core.Api.Types open FSharp.CloudEdge.Management.Storage let createDatabase (storage: StorageClient) accountId = task { match! storage.D1CreateDatabase(accountId, D1CreateDatabasePayload.Create "guestbook") with | D1CreateDatabase.OK payload -> let databaseId = string payload.result["uuid"] printfn "D1 database guestbook: %s" databaseId return Some databaseId | D1CreateDatabase.Status4XX(status, failure) -> for error in failure.errors do printfn "HTTP %d: %s" status error.message return None } ```

Needs a StorageClient and your account ID, as in the Control Plane API Client.

## Photo Bucket Request records have a `Create` method that takes the required fields and sets the rest to `None`. A `with` expression then sets one of those fields. Here it adds a location hint for Western North America, which Cloudflare treats as best effort. ```fsharp open FSharp.CloudEdge.Core.Api.Types open FSharp.CloudEdge.Management.Storage let createBucket (storage: StorageClient) accountId = task { let bucket = { R2CreateBucketPayload.Create "photos" with locationHint = Some r2_bucket_location.Wnam } match! storage.R2CreateBucket(accountId, bucket) with | R2CreateBucket.OK payload -> printfn "R2 bucket %O in %O" payload.result["name"] payload.result["location"] return true | R2CreateBucket.Status4XX(status, failure) -> for error in failure.errors do printfn "HTTP %d: %s" status error.message return false } ``` ## Session Store The new namespace's `id` is the `namespace_id` of a `kv_namespace` binding. Cloudflare returns a 400 when the account already has a namespace with the same title, so a second run of `createNamespace` ends in the `Status4XX` case. ```fsharp open FSharp.CloudEdge.Core.Api.Types open FSharp.CloudEdge.Management.Storage let createNamespace (storage: StorageClient) accountId = task { let body = workers_u002D_kv_create_namespace_body.Create "sessions" match! storage.WorkersKvNamespaceCreateANamespace(accountId, body) with | WorkersKvNamespaceCreateANamespace.OK payload -> let namespaceId = payload.result |> Option.map (fun created -> created.id) printfn "KV namespace sessions: %A" namespaceId return namespaceId | WorkersKvNamespaceCreateANamespace.Status4XX(status, failure) -> for error in failure.errors do printfn "HTTP %d: %s" status error.message return None } ``` ## Order Queue `QueuesCreate` takes its body as an optional argument, so the call passes it by name. The consumer example on [Worker Upload](worker-upload.md) takes the `queue_id` from this result. ```fsharp open FSharp.CloudEdge.Core.Api.Types open FSharp.CloudEdge.Management.Compute let createQueue (compute: ComputeClient) accountId = task { match! compute.QueuesCreate(accountId, body = QueuesCreatePayload.Create "orders") with | QueuesCreate.OK payload -> let queueId = payload.result |> Option.bind (fun queue -> queue.queue_id) printfn "Queue orders: %A" queueId return queueId | QueuesCreate.Status4XX(status, failure) -> for error in Option.defaultValue [] failure.errors do printfn "HTTP %d: %s" status error.message return None } ``` ## Bucket List `R2ListBuckets` returns its result as a `JsonObject` with a `buckets` array, and the loop reads each bucket's `name` and `creation_date`. ```fsharp open FSharp.CloudEdge.Core.Api.Types open FSharp.CloudEdge.Management.Storage let listBuckets (storage: StorageClient) accountId = task { match! storage.R2ListBuckets(accountId) with | R2ListBuckets.OK payload -> for bucket in payload.result["buckets"].AsArray() do printfn "%O created %O" bucket["name"] bucket["creation_date"] | R2ListBuckets.Status4XX(status, _) -> printfn "HTTP %d" status } ``` ## Resource Inventory `WorkersKvNamespaceListNamespaces` and `QueuesList` return F# records. Optional schema fields are `option` values, such as a queue's `queue_id`. ```fsharp open FSharp.CloudEdge.Core.Api.Types open FSharp.CloudEdge.Management.Compute open FSharp.CloudEdge.Management.Storage let listNamespaces (storage: StorageClient) accountId = task { match! storage.WorkersKvNamespaceListNamespaces(accountId) with | WorkersKvNamespaceListNamespaces.OK payload -> for kv in Option.defaultValue [] payload.result do printfn "KV %s %s" kv.id kv.title | WorkersKvNamespaceListNamespaces.Status4XX(status, _) -> printfn "HTTP %d" status } let listQueues (compute: ComputeClient) accountId = task { match! compute.QueuesList(accountId) with | QueuesList.OK payload -> for queue in Option.defaultValue [] payload.result do printfn "Queue %A %A" queue.queue_id queue.queue_name | QueuesList.Status4XX(status, _) -> printfn "HTTP %d" status } ``` ## Library Table | Library | Operations | What it covers | | --- | ---: | --- | | `Management.Storage` | 153 | R2, D1, Workers KV, Vectorize, Hyperdrive | | `Management.Compute` | 304 | Workers, Queues, Workflows, Pages | ## Related Pages
Worker UploadBindings for these resources StorageKV, R2 and D1 in a Worker Background WorkQueue messages in a Worker Control PlaneClients and the API token
## NuGet packages [Core.Api 0.1.0](https://www.nuget.org/packages/FSharp.CloudEdge.Core.Api/0.1.0), [Management.Compute 0.1.0](https://www.nuget.org/packages/FSharp.CloudEdge.Management.Compute/0.1.0), [Management.Storage 0.1.0](https://www.nuget.org/packages/FSharp.CloudEdge.Management.Storage/0.1.0). See [installation and release availability](../../guide/packages.md).