ROR API performance issues (resolved for now)

In the past 2 weeks, we've seen significantly increased traffic in the ROR API due to what appears to be about 50 instances of the same application sending a high volume of repetitive requests to the the affiliation matching endpoint constantly for several days at a time. All IPs are AWS eu-west-1, and the client is Axios (node.js). We saw this activity from 13-18 Sep and again beginning 25 Sep. 

This has caused some degraded performance in the API, but we've taken some temporary measures to mitigate this behavior and performance should now be back to normal. If this might be your application(s), please contact [email protected] to discuss.

As always, ROR is a shared community resource provided at no cost to users, so please be considerate of other users when configuring applications that use the ROR API.  This is also a good time to remind you that we have a proposal about API client identification and rate limiting open through 4 Oct! Please add your feedback to the doc as comments: https://docs.google.com/document/d/1_4e75i8RWl82Iz4CwMpTSQ2ki3H4xRKkx40-N-ih9nA

Cheers,

Liz