# Plan Bundles

Every Wokku box is batteries-included: five shared engines are available at no extra cost, and you pick which ones to attach at box creation. There is no add-on store and no separate addon invoices. Paid plans additionally include a quota of dedicated engine slots for workloads that have outgrown shared tenancy.

Plan matrix

“3 DB slots” covers all three DB engines combined (Postgres, MySQL, MongoDB). A Pro workspace with 2 dedicated Postgres DBs + 1 dedicated MySQL has used all 3 DB slots.

Shared add-ons: pick per box

The five shared engines are:

Each engine you enable on a box gets its own isolated tenant — a dedicated role + database for Postgres, a key prefix for Redis and Memcached, a virtual host for RabbitMQ, an index namespace for Meilisearch. Tenants are isolated; one box cannot read another’s data.

You can enable or disable engines after creation from box → Settings → Add-ons. Disabled engines keep data for 24 hours before the tenant is cleaned up.

Dedicated engine slots

When a shared tenant is no longer enough — you’ve hit the size cap, need full admin access, or want to share a DB across multiple boxes — you can upgrade to a dedicated single-tenant container. Upgrade consumes one slot from your quota.

All three DB engines (Postgres, MySQL, MongoDB) share the same 3-slot pool. Redis has its own 3-slot pool.

Size follows the box: a medium box gets a medium dedicated engine. To get a bigger engine, resize the box first.

See Dedicated upgrade for the full migration flow.

Extra boxes (Team plan only)

Only Team plan workspaces can buy extra boxes beyond the included 3 sleeping + 3 large:

• Large — $10 / Rp 150k per box per month
• X-Large — $20 / Rp 300k per box per month

Free, Solo, and Pro pay only the flat plan price and cannot add extra boxes.

Workspace MinIO

MinIO is workspace-level object storage — not a per-box add-on. Every paid plan includes a shared bucket; the size scales with plan. Credentials are in Workspace → Storage.

MinIO is in early access. Bucket creation and the S3-compatible API are stable; the dashboard object browser ships in a later release.

Why bundle them?

The per-add-on billing model used by Heroku’s marketplace or Vercel’s integrations works, but it creates friction:

• You end up with 5+ line items on every invoice.
• Cost compounds: a small DB + small Redis + small search engine can be 3× your compute cost.
• You spend time picking tiers across multiple providers, and they don’t scale together.

Wokku’s model: compute + cache + DB + search + queue, all included in the plan price and sized as a coherent set. The Pro plan’s medium boxes are matched to medium dedicated engines — you don’t hit one cap before the other.

Exceeding a shared cap

If a shared engine tenant hits its size limit:

• Postgres — writes are blocked with a 24-hour grace period; old rows stay readable. The SharedDatabaseQuotaJob warns at 80%.
• Redis — LRU eviction kicks in (oldest keys dropped first), or NOEVICTION returns errors, depending on your maxmemory-policy.
• Memcached — entries are evicted normally (Memcached is always LRU).
• RabbitMQ — new messages are rejected when the vhost queue depth limit is reached.
• Meilisearch — indexing returns errors when the index size limit is reached; existing indexed data remains searchable.

The right fix is usually upgrading to a dedicated engine (frees the shared slot) or moving to a higher plan (bigger dedicated size).

Self-hosted

Community Edition has no bundle system — you manage Postgres, Redis, and other engines directly via your Dokku setup. No caps, no quotas.

See also

• Plans & Pricing — what each plan costs and how seats work
• Supported engines — all 5 shared + 4 dedicated engines
• Create & Link add-ons — enable shared add-ons, CLI/API/MCP reference
• Dedicated upgrade — upgrade flow, quota, sizing, rollback
• Usage — live utilization tiles per workspace