Help

API Ping endpoint for status checks?

Topic Labels: API
Solved
Jump to Solution
1515 1
cancel
Showing results for 
Search instead for 
Did you mean: 
Adrian
4 - Data Explorer
4 - Data Explorer

Hello!

I help run https://vacfind.org, a COVID vaccine information site that stores links to helpful websites in Airtable. Given the recent Airtable downtime, I was wondering if it would be possible to have a simple /ping endpoint (or similar) added to the API so that I can configure our caching server to use this endpoint to check whether Airtable is up and if not, configure itself to serve requests from the cache.

Does something like this already exist in the Airtable API that I’m not already seeing? ideally id like to not have to configure auth to make a request just to figure out if the Airtable API is up (200 OK) or not (503 Service Unavailable).

Thanks,
Adrian

1 Solution

Accepted Solutions
Bill_French
17 - Neptune
17 - Neptune

Hi @Adrian, and welcome to the community!

There’s no easy way to detect if an API is alive because a PING is insufficient evidence of call-worthiness and anything short of full authentication is incapable of validating if their authentication layer is even running. Ergo, your only reasonable method is a few carefully constructed lightweight calls from time-to-time that perform general monitoring of the API. And even this is not ideal because 5 minutes ago doesn’t really matter when it’s time to do actual work. As such, you don’t monitor - you build solutions that are smart and roll with the outages in a graceful manner.

See Solution in Thread

1 Reply 1
Bill_French
17 - Neptune
17 - Neptune

Hi @Adrian, and welcome to the community!

There’s no easy way to detect if an API is alive because a PING is insufficient evidence of call-worthiness and anything short of full authentication is incapable of validating if their authentication layer is even running. Ergo, your only reasonable method is a few carefully constructed lightweight calls from time-to-time that perform general monitoring of the API. And even this is not ideal because 5 minutes ago doesn’t really matter when it’s time to do actual work. As such, you don’t monitor - you build solutions that are smart and roll with the outages in a graceful manner.