Two eyes are better than one eye

OpenStack produces tons of log files in a real production environment. It becomes harder for a cloud operating team to analyze and parse them by extracting data in each file using a combination of tail, grep, and perl tools. The more hosts you build, the more logs you have to manage. Growing a few paces should be companioned by a serious trace keeper. To overcome such a challenge, your log environment must evolve to become centralized. A good option can be by starting flowing logs in a dedicated rsyslog server. You may put in so much data and your log server start starving for larger storage capacity. Furthermore, archiving the former data will not be handy when you need to extract information for a particular context. Additionally, correlating logs data having different format (taking into consideration RabbitMQ and MySQL logs) with generated events might be even impossible. So...