IBM/OpenWhisk函数触发源及Connections、Polling适用范围咨询

Let's break down your two questions clearly, based on how OpenWhisk and IBM Cloud Functions work:

1. Can IBM/OpenWhisk functions be triggered by triggers from any source? #

Absolutely—you have multiple ways to hook up triggers from any external system or custom source, as long as you can establish a communication path:

• Webhooks: The simplest option for systems that support HTTP POST requests. You can create a webhook trigger in IBM Cloud Functions, then point your external service (whether it's a SaaS tool, self-hosted app, or custom script) to send events to that trigger's endpoint. This will fire your associated function immediately.
• Custom feeds: If your source doesn't support webhooks (like a legacy system or a service that only exposes a read-only API), you can build a custom feed using OpenWhisk's feed framework. This lets you write code that monitors the source for changes and triggers your function when events occur—this works with pretty much any system you can interact with programmatically.
• Direct API calls: Even without a trigger, you can call your OpenWhisk function directly via its REST API from any source that can make HTTP requests. This skips the trigger layer entirely if you don't need the event routing features of triggers.

2. Are Connections and Polling feeds limited to the IBM Cloud ecosystem, or can they work with arbitrary services (like self-hosted databases)? #

Let's split this into the two feed types:

• Polling feeds: These are fully platform-agnostic. Polling feeds work by periodically checking a specified endpoint or resource for updates (like new records, changed files, or status updates). You can configure them to poll a self-hosted database's API, a custom REST service running on your own servers, or even a third-party tool's endpoint—there's no restriction to IBM services. You just need to define the polling interval, the target URL, and how to detect changes (e.g., comparing response timestamps or record counts).
• Connections feeds: These are purpose-built for IBM's ecosystem. The pre-built Connections feeds are designed to integrate seamlessly with specific IBM Cloud services (like IBM Cloud Object Storage, Watson services, or Event Streams). While you could technically build a custom Connections feed for a non-IBM service (using the same framework), the out-of-the-box options are limited to IBM's platform. For self-hosted services, a custom polling feed or webhook is usually the more straightforward approach.

内容的提问来源于stack exchange，提问作者nymvno