Blog & News

Weekly Report

API.expert Weekly API Analysis: Twitter OAuth

Apr 23, 20213 min read
PC

Written by

Paul Cray

Co-Founder

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

This Week’s API Highlights

  • Google in the Social Networks API category takes the overall title this week with a CASC score of 9.68.
  • Department of Justice in the US Government API category takes the overall title again with a median latency of 47 ms, the same as last week
  • Environmental Systems Research Institute in the US Government API category takes the overall bottom spot this week with a CASC score of 6.52.
  • We take a good, hard look at what went wrong with Twitter API this week.

Something of interest

Last week, Twitter was riding high with a first-class API CASC score of 9.74, first in the Social Networks API category. This week it is languishing in the doldrums with a CASC score of just 8.27, placing it second from last. What went wrong? api.expert apimetrics api analysis Twitter OAuth 1 Last week, it had a 100% pass rate and just 0.30% outliers. This week has a 99.27% pass rate, very far from 5 Nines Telco strength expected of such a crucial piece of web ecosystem, and 4.98% outliers. It is also nearly 17% slower. api.expert apimetrics api analysis Twitter OAuth 2api.expert apimetrics api analysis Twitter OAuth 2 A glance at the API Insights table tells us immediately which was the recalcitrant endpoint: Twitter OAuth 1.0 request token. Above we see the performance affecting outliers for the week beginning 5 April 2021. And below we see them for the week beginning 12 April 2021. The display has lit up like a Christmas tree! api.expert apimetrics api analysis Twitter OAuth 3 There was a continuous outage that lasted from 01:11 UTC on 17 April 2021 to 05:15 on 17 April 2021. That’s a major outage, more than four hours. Anyone trying to get a token during that period would have been out of luck. We also saw quite a few outliers later that day and continuing to 18 April. Not everybody using your service might be calling that endpoint during that time, so only some of your users are going to be adversely affected. But this is why you should be actively monitoring your APIs. Active monitoring provides a heartbeat. You can see immediately when your endpoint went down and when it came back. You are not reliant on your users telling you that something has gone wrong. You know straight away and can start fixing things. A glance at the HTML returned by the HTTP status code 503 Service Temporarily Unavailable response tells us that this wasn’t a planned outage, so let’s hope that not too many people’s days were spoiled!

API Analysis: Tops in Overall Quality

Week Ending 19 April 2021

Category

Organization

CASC score

Corporate Infrastructure

Slack

9.57

COVID-19

WHO

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