A feature flag, a preferred theme, a temporary announcement and an old-page redirect all need a small piece of stored data. Workers KV lets a Cloudflare Worker look up a value by its key without requiring you to run a database server. It is useful for data that is read often and changed less frequently.
This course introduces each idea through a working application task. You will first keep local and cloud data separate, then read structured settings, handle missing values, set expiration and traverse paginated results. You will also learn why a recently updated value may not be visible to every reader immediately—and how to keep optional display settings usable during that delay.
The course contains 5 guided labs and 1 independent notice-repair challenge. Guided Labs introduce each capability before the independent challenge.
What You Will Learn
- Create a Feature Flag Store: turn a local banner on while the cloud setting stays off, then read both through a Worker binding.
- Serve Account Preferences: give two fictional accounts different JSON settings and use clear defaults for missing or damaged records.
- Expire Temporary Notices: stop displaying a notice at its application deadline and separately observe KV expiration.
- Import a Redirect Catalog: validate and bulk import a small dataset, read every page, and retire selected keys without losing unrelated data.
- Handle Delayed Configuration Updates: test older and newer settings safely, then observe a real remote update without requiring a stale read to occur.
- Missing Maintenance Notice: repair a namespace selection and notice-handling problem while preserving separate live and preview data.
Who This Course Is For
Workers developers who need small settings, announcements and redirects without managing a database server.
Prerequisites: Complete Cloudflare Workers for Beginners, especially Connect LabEx to Your Cloudflare Account. That lab explains the VM terminal, Wrangler device authorization, account confirmation and account-ID configuration. You should also recognize a JavaScript Worker handler and know how to deploy and remove a Worker.
Learning environment: Use your own Cloudflare learning account. Each lab starts in a fresh LabEx VM and creates independent disposable resources; a previous VM's login or namespace is not reused. The supplied small synthetic datasets need no purchased domain or paid upgrade. Account allowances still apply, so keep the datasets small and complete the documented cleanup.
These examples use KV for display settings and other read-heavy data. They do not use a potentially stale value to decide immediate access revocation or financial transactions. You will check actual responses and cloud resources, remove only your lab resources, and log out before finishing.
Frequently Asked Questions
What kinds of data suit KV?
Display settings and other read-heavy data that changes less frequently. These exercises do not use potentially stale values for immediate access revocation or financial transactions.
Will every reader immediately see an update?
No. The course tests older and newer settings safely, then observes a real remote update without requiring a stale read to occur.
Is an application deadline the same as KV expiration?
No. The notice lab stops displaying expired content at its application deadline and separately observes storage expiration. Missing or damaged records need clear handling.
Do local and cloud namespaces share data?
They are separate targets. You verify the selected namespace and responses, preserve unrelated data during repair and clean up only your own lab resources before logging out.
