Good question @Anil Sagar. I think webhooks enable real or near real-time integration of decoupled APIs/systems that in the past it was just too difficult.That being said, when we're talking about webhooks, we're just talking about event/message-driven architecture (EDA), which requires careful consideration regarding reliability and complexity. Here some thoughts:
- How critical is the operation of webhooks for your business? Mission critical? Then you need to think more about providing the right infrastructure to support expected reliability. i.e. implementing messaging queues. If the business can afford the eventual loss of messages, maybe putting more safeguards are just overkill, perhaps not. It is important to calculate these risks to measure the level of effort to implement this functionality. Fortunately, all the infrastructure required to achieve desired reliability is available from most cloud providers.
- If reliability is a concern, on the implementation side, you'll also need to consider what happens when webhooks fail to deliver a message to a consumer API/system?
- Will it support retries? If so, how often? How long?
- Or will you allow external systems to reply requests?
- How will you scale this infrastructure?
- Make sure webhooks don't impact the regular flow of your APIs. Make sure it doesn't impact latency, for instance, ensure that webhooks execute in a fire-and-forget fashion.
- On usability side
- How will you manage subscribers to webhooks? Self-Service?
For more details about the implementation of a webhook facade, we created a sample Node.js webhook API Proxy that sits on top of the Management API. It also implements some pub/sub features.
Hope it helps!
LinkedIn
Twitter
