A super efficient Amazon SQS thread based message processor for Ruby. This project is in MAINTENANCE MODE. Reach me out on Slack link on the description if you want to become a new maintainer.
Other
2.06k
stars
280
forks
source link
Shutdown when workers encounter unrecoverable errors #773
Postgres database managed by AWS RDS, with password management configured
Database credentials configured by loading an AWS Secret into the environment
Roughly every 7 days RDS rotates the password and updates the associated secret. This cannot automatically update the environment of the running ECS instances.
Eventually ActiveRecord attempts to create a new DB connection and fails raising an ActiveRecord::NoDatabaseError. I believe is being caught by middleware/server/exponential_backoff_retry.rb
What I really want is for the shoryuken process to be shutdown when this error is encountered, as ECS will then restart the container with refreshed environment properties.
however reading through the code and comments on other issues it looks like this is not possible. Any suggestions?
Background:
Roughly every 7 days RDS rotates the password and updates the associated secret. This cannot automatically update the environment of the running ECS instances.
Eventually ActiveRecord attempts to create a new DB connection and fails raising an ActiveRecord::NoDatabaseError. I believe is being caught by middleware/server/exponential_backoff_retry.rb
What I really want is for the shoryuken process to be shutdown when this error is encountered, as ECS will then restart the container with refreshed environment properties.
however reading through the code and comments on other issues it looks like this is not possible. Any suggestions?