Weekly Report

API Ratings: Nov 10-17 | Foursquare Outage

Nov 19, 20203 min read

Written by

Paul Cray

Co-Founder

Paul co-founded APIContext (formerly APImetrics), building the synthetic monitoring and conformance testing platform from its earliest days.

API Performance

We look at over 100 APIs and pull all the metrics together to give you the API ratings for an organization’s APIs in a particular category with our patented CASC score.

Something of interest

Foursquare had a bad week last week claiming the wooden spoon in both the Search and Social Network categories. What went wrong? Foursquare had an outage on 12 November 2020 that started just after 15:00 UTC and lasted until after 21:30 UTC. It was an intermittent outage, so not all calls were failed during that time and there were a handful of fails afterward (and before), but the main outage lasted about 6 and a half hours. This kind of thing can really ruin your Thursday. You really don’t want to be staring at a screen of red like this. api-ratings-foursquare-outage-1 What’s interesting is that we see a whole plethora of different errors:

HTTP Status Code 500 Internal server error "Foursquare servers are experiencing problems. Please retry and check status.foursquare.com for updates.”
HTTP Status Code 502 Bad Gateway
HTTP Status Code 503 No healthy backends
HTTP Status Code 503 backend read error
HTTP Status Code 503 All backends failed or unhealthy
HTTP Status Code 504 Gateway Time-out

These definitely aren’t the sorts of messages that you want to be seeing! That’s why it’s so vital to use an active API performance and monitoring tool. It lets you dig down into what went wrong in a way that gateway logs alone don’t let you do. api-ratings-foursquare-outage-2 Here you can see the massive spike in latency on 12 November. The Foursquare API was pretty stable through late October and into November before things fell apart. The outage lights up like the proverbial Christmas tree on the Performance Impacting Outliers chart. You want to see when a problem has arisen – and you can do that with an active monitoring system – but you can then use the test call data you have obtained to forensically analyze the outage. This can help you spot outages in the future or put in place measures to prevent them in the first place.

Top API ratings: overall quality

Week Ending 16 November 2020

Category

Organization

CASC score

Corporate Infrastructure

GitHub

9.46

COVID-19

covid19api.com

See what your APIs look like from the outside.

APIContext gives engineering, product, and customer success teams a shared view of API reliability, conformance, and customer impact — without rebuilding dashboards.

Start free