How often - and under what circumstances - does the Microgateway "pull" from Edge?

[willwitman · Accepted solution]

@Justin Moses

That's a good question. Here's a list -- hopefully others( @prabhat ) will help point out what's missing or wrong so we can get a nice clean list.

(This answer is mainly specific to Edge Micro 2.0.)

• edgemicro configure — Deploys a proxy used for authentication, stores key information in a secure store, gets back keys needed to start Edge Micro. You run it once. The config information is cached in the ~/.edgemicro directory.
• edgemicro start - Pulls down proxy and product information from your Edge org (used for API key and OAuth authentication and quota enforcement). This info is cached in the ~/.edgemicro directory.
• API key auth — If you use API key auth, then the key is validated on Edge, and then cached so further API calls don’t have to re-authenticate with Edge.
• OAuth token — When you get an OAuth Bearer token for OAuth security, you need to obtain that token from Edge, but you only need to do that once using a CLI command.
• Other less-often used management commands communicate with Edge and return data, like edgemicro cert, genkeys, and token.
• Once started, Edge Microgateway communicates with Edge whenever it asynchronously pushes analytics data to Edge. I’m not sure of the time between analytics updates -- maybe every few hundred milliseconds.

[#2]

• Once started, Edge Microgateway communicates with Edge whenever it asynchronously pushes analytics data to Edge. I’m not sure of the time between analytics updates -- maybe every few hundred milliseconds.

You can trace the proxy to see. On our systems it seems to communicate in batches of ~ 100 messages. This might even be configurable.

[#3]

We really need a better way of seamlessly integrating changes to products w/ edge micro configs. The product configurations are NOT an it function in my shop - they are a business function. And ... thats a problem right now.

[prabhat]

Yes, fair point.