Messages should be cached in process.

As a starting point I would suggest the TTL is 1min. Assuming 1request/second and only 1 instance this gives us a hit rate of around (60/59)=98% and reduces our number of hits to the API by a factor of 60. On the other hand 1 min is an acceptable delay to get new messages.

After talking on IRC with the serviceops people it sound like only having 1 instance is likely but we should be prepared for it to scale up slightly if needed with load. We would expect ca. 1-3 instances (probably 1) and not 10-50.

Notes

• there is a axios cache adapter
• there is also axios-extensions
• https://github.com/axios/axios/issues/31 has some simple library-free suggestions

Technical notes:

• can be enabled on a repo/action level (not only globally)
• TTL can be configured on repo/action level, config coming from environment variable
• ideally not touching existing repos but added transparently
• independently of implementation, the cache data source should only maintain a limited number of items - hard coding this to 1000 should be a safe bet
• do not implement your own cache solution! code for wiring up existing libraries (e.g. https://www.npmjs.com/package/lru-cache) should suffice