I am super excited to introduce guest blogger, Ana Krivokapić, who ran over to OpenStack Days Poland to represent Red Hat as the TripleO RDO OpenStack guru in residence. Thanks so much to Ana for attending, answering any technical TripleO RDO OpenStack questions sent her way, and reporting back on her experience.
A few weeks ago I had a privilege of attending the OpenStack Days Poland in lovely Warsaw, Poland. It was the first official OpenStack Days event in Poland, and only the fifth OpenStack event ever organized in Poland.
The sessions were a healthy mix of high level overview-type talks and talks that went deeper into technical or business details of a particular aspect / project / component of OpenStack. The attendees were a solid mix of OpenStack developers, experienced users / operators of OpenStack, and potential users who came satisfy their curiosity about OpenStack and learn more.
The event was sponsored by a number of companies that had booths where attendees could talk to the company representatives / recruiters, get some swag, or even play video games. For the stats / fun facts enthusiasts out there: OpenStack Days Poland had 268 attendees, 24 talks were given by 34 speakers, and together we drank 86 liters of coffee over the course of the day.
The first keynote, “The Future of OpenStack – Trends and Insights from the User Survey”, was given by Heidi Tretheway and Heidi Bretz from the OpenStack Foundation. They gave an insightful look into the latest OpenStack User Survey conducted by the OpenStack Foundation. The trends are showing the OpenStack user base is maturing in every way: we now have companies of every size and from every industry using OpenStack in their production deployments. As opposed to the early stages when mostly research companies were OpenStack users, we are now seeing enterprise users as well as small-medium size companies, which is a great testimony of OpenStack’s maturity. More insights from the User Survey can be found at https://www.openstack.org/user-stories and https://www.openstack.org/analytics.
The second keynote was “Bringing Cloud Native Innovation to the Enterprise” by Rob McMahon from Red Hat. He talked about how Red Hat customers use OpenStack: 42% use it in production deployments, 66% use some sort of a PaaS on top of it and 58% use containers on top of it. Use cases include building a private cloud (NASA; fun fact: Mars Rover runs on OpenStack!), massive transactions exchange (Paddy Power Betfair), massive NFV cloud (Verizon), public cloud provider (RackSpace) and DevOps (Produban). Rob offered some very interesting insights, such as that the culture and process are as important as technology, as well as that the CIOs now want their IT departments to evolve from cost centers to innovation centers.
The third keynote, “Contributing to OpenStack – above divisions”, was given by Michał Pryć from Red Hat and Michał Jura from Suse. They talked about the first steps needed to become an OpenStack contributor and explained how the OpenStack community operates.
Tomasz Trębski and Artur Basiak of Fujitsu gave a talk called “Monitoring of Openstack clouds”. They presented the CMM (Cloud Monitoring Manager) tool developed by Fujitsu engineers. CMM is a single tool to monitor OpenStack deployments while ensuring stability and availability. It is based on the OpenStack project Monasca, which provides monitoring- and logging-as-a-service. CMM has various metrics, alarms, notifications, a CLI and a GUI (integrated into Horizon).
“Dragonflow – Doing Cloud Networking to Scale” by Omer Anson, the PTL of the Dragonflow project, introduced us to Dragonflow – a fully distributed backend for Neutron. Omer demonstrated that Dragonflow is built to scale and it scales well (e.g. creating 50 subnets in less than a second).
The last talk I attended was “Deploying OpenStack with TripleO: An in-depth look” by Gregory Charot from Red Hat, who gave a high-level overview of the TripleO project, which uses OpenStack services to deploy OpenStack. He examined the TripleO lifecycle (planning, deployment, operations, and tooling), as well as the steps in the deployment flow (get hardware, rack and wire it, identify management node, install undercloud, identify overcloud nodes, benchmark hardware, define networking, register and introspect nodes, tag / assign nodes, and kick off overcloud deployment). Gregory explained how to use Heat templates to customize the deployment, how to configure networking, what the composable roles are, which Ceph integration options are supported, how to scale the overcloud up and down, and how to perform minor updates and major upgrades.
Overall, I was very impressed with the organization of the OpenStack Days Poland as well as the quality of the talks themselves. I did have two small objections, though. First, based on the schedule itself, it was impossible to know in what language, English or Polish, a particular talk would be given. Some talks had the title and summary written in English on the schedule but the talks itself was in Polish. Second, the OpenStack Days Poland website was in Polish only, so it was a bit tricky for a non-Polish speaker to register. Online translation services to the rescue! Even so, these were only minor hiccups and my overall impression of the event was very positive.