The complex architecture, design, and difficulties while troubleshooting amplifies the effort in debugging a problem with an OpenStack environment. This can give administrators and support associates sleepless nights if OpenStack native and supporting components are not configured properly and tuned for optimum performance, especially with large deployments that involve high availability and load balancing.
We, being support associates dedicated for OpenStack, along with system administrators have gone through a lot of sleepless nights to chase problems like this. Throughout this session we will discuss some of those problems in depth, how we learned to troubleshoot them, what the root cause was, and how we arrived at the right solution. Some of these problems are related with scaling load balancers that sit in front of API services, scaling rabbitmq message bus, scaling connections to MariaDB + Galera, and many more.
Attendees will have good understanding on some of the most commonly known issues with openstack that can hit their environment. This will help system administrators, architects and openstack users to be aware of these problems, scan their openstack deployment, proactively identify possibility of hitting them and apply these solutions before they are hit.