Managing logging in your OpenStack environment

Given an OpenStack deployment where services are varied and distributed across many hosts, creating an efficient and usable logging solution can present quite a few challenges. From an Operators perspective, not having useful or easily accessible logs can massively hamper issue resolution and health checking efficiency, but what can we do to ensure the logs are as useful and accessible as possible?