how an application in nubis would use fluent to push it's logs out so that they could be accessed by Opsec?
how the fluent aggregation works (pointer to IAM role it uses, pointer to s3 buckets it uses, pointer to security group rules that allow clients to talk to aggregator, etc)
In our conversation a few weeks back you described this (which I took notes about), but I'd like to get the official story (what you've actually implemented)
A Fluentd collector will be deployed in both the staging and production VPC for each product ( http://www.fluentd.org/ )
This collector will have an IAM role granting it write access to an S3 bucket
It will have a security group granting Shared Services systems to connect inbound to it
It will announce its logging service to consul
Fluentd agents will be deployed on all instances
The agents read on disk local log files
Parse them to turn them into structured data
Send the structured data to Fluentd collector over TCP in JSON
Can you link to or write up a description of
In our conversation a few weeks back you described this (which I took notes about), but I'd like to get the official story (what you've actually implemented)