Failover Provider

Generic provider failover wrapper for retrying sync and async calls across provider candidates

Failover Provider wraps multiple provider-like objects behind one proxy so your integration can retry failed calls against the next candidate without rewriting each call site. Use this page to review the Configuration and API Reference for the released @tetherto/wdk-failover-provider surface.

Features

Generic provider wrapper: FailoverProvider\<T\> accepts any provider-like object type and returns a proxied T from initialize().
Configurable retries: Set retries to control how many additional attempts happen after the first failure.
Retry predicate: Use shouldRetryOn(error) to decide which errors should advance to the next provider.
Sync and async failover: The runtime retries both synchronous throws and rejected promises.
Zero runtime dependencies: The published package ships a single default export with no runtime dependency tree.

Why this matters

You can keep one provider-shaped integration surface while rotating between browser, JSON-RPC, or custom provider implementations.
You can narrow failover behavior to transient errors instead of retrying every exception.
You can add redundancy to read-heavy or submission-heavy flows without building your own retry proxy.