Requests

Experimental support out now: https://www.inngest.com/docs/self-hosting. Run the open source Inngest on your own infrastructure with minimal dependencies. Postgres support coming soon.

Provide an outbound connection from the SDK to the Dev Server and Cloud. Allowing long lived connections to handle long running jobs which can exceed HTTP request timeouts. This also allows users to run the SDK within their network without opening ports.

Allow users to configure alerts for automatic notifications when functions fail

Forward database change events locally to Inngest to trigger functions

When deploying crons, having them run in each branch environment can make testing more difficult and increase function calls unnecessarily. Users can still manually invoke crons in Branch Envs.

Currently under consideration.

Steps should be able to override retry policies and set their own concurrency keys

Allow passing in regions, eg: eu, within HTTP headers when sending events, ensuring that all data is parsed, stored, and executed within the specified region.

Tools like superjson are common in the JS ecosystem, allowing users to maintain objects such as dates across network boundaries. The SDK can support these patterns by allowing a "transformer" to serialize/deserialize data passing to/from Inngest.

Add the ability to set a timestamp in the past for waitForEvent to retroactively look backwards for an event. This is useful for race conditions and waitForEvent.

Forward database change events to Inngest to trigger functions (change data capture)

Ability to see webhook requests sent to Inngest. It will help users create and test transform functions and debug webhook issues, such as mismatch of event format.

Write Inngest functions in PHP and serve them using your favorite frameworks like Laravel.

Allow configuration of step timeouts on a per-step basis, cancelling work if the timeout is reached before steps finish.

When rerunning a function, start the new run from an existing step in a previous run.

Invoke an Inngest function outside of the scope of a function handler and await it's result.

Improving the DX os webhooks to make it easier to write, test, and debug issues with webhook transforms.

Build an SDK that allows defining Inngest functions using Ruby

Define triggers (events or crons) via an API rather than in a function definition

Add the ability to share throttle configuration across functions with scopes.

Store function failures and allow full text search (eg. for filtering, which lets us replay specific errors). Categorize errors, eg. for function timeouts, and show smart tips to resolve.

Bulk reprocess past events

Enable users to select webhooks within their Inngest Cloud account that they want to temporarily forward to a running Dev Server. Useful for testing webhook functions during local dev.

Expose low, normal, and high priorities for functions. Potentially allow replays to have their own priorities (note that if a function has high priority and a replay has high priority, this will be a noop).

Currently under consideration.

New roles for varying levels of access within an Inngest account.

A pre-build Github action that will trigger an App sync via Inngest REST API. This should provide a more seamless syncing experience similar to the Vercel integration for all users.

Currently under consideration.

Add filtering by app, sorting and a filter text input

A user testing function that use "sleep" or "sleepUntil" should be able to click a ▶️ or similar button that forces the function to continue and complete the sleep faster. This is key to testing long sleeps.

Currently under consideration.

Add a tool which allows for continual linear subscriptions to an event stream, eg: step.subscribe("event/name", { if: "xyz == 123", timeout: "10m" }, (acc, async, ctx) => {}). This lets us listen for all events of a type within a function.

Use an SDK to add key-value tags to a given function run enabling future search and filtering.

Improved environments overview page for creating and managing environments including branch environments.

Users are able to use autocomplete when they type their payloads, based on the schema defined

Log skipped runs along side events that did not trigger function runs due to flow control configuration like rate limit, debounce, idempotentcy.

Allow webhook transforms to customize the response status code, headers, and body

Add ability to manually archive events

Forward metrics and other Inngest system data to New Relic.

Currently under consideration.

Add CEL expression filtering to bulk cancellation and better integrate into the Cloud dashboard experience.

Currently under consideration.

When a user sees a failing run, they

UX improvements to archived environments

Provide a drop down with typeahead filter global runs views within Cloud + Dev Server

Adding the "connect" API for long running connections to the Python SDK

A flow control option to ensure that if another run will not be triggered if there is a run that is actively running with the same key.

Allow exporting metrics to Datadog and Prometheus