tree 787f148dcbb90ea01776c5a4a74607283fbc2eea
parent 32fb8bbd346b7e3755f59688bcb20e35eb790bc0
author Jason Peterson <jasonpet@us.ibm.com> 1495743338 -0400
committer Justin Berstler <bjustin@us.ibm.com> 1495743338 -0400

Implement provider redundancy (#100)

This change introduces redundancy in the cloudant feed provider, allowing multiple instances to be deployed simultaneously such that if any one instance stops working, the remaining instances can be activated to take over the work of the failed instance.

This is accomplished by changing the feed actions to write directly to the trigger DB, rather than talking directly to the feed provider through REST endpoints. In this way, provider instances become aware of new and deleted triggers by monitoring the DB changes feed.