Using Ansible Tower To Automate Baremetal Benchmarking - AnsibleFest Atlanta 2019
At AMD we manufacture semiconductors, specifically CPUs and GPUs. As you might expect, we have to collect a LOT of performance data on them during development (and after). This is a case study on how AMD implemented Ansible, Red Hat Ansible Tower, and RedHat OSP to create a bare metal deployment service to allow for automated benchmark runs. It talks about the playbooks for both OS deployment and benchmark deployment and execution.
A Telco Story of OpenStack Success - OpenStack Summit Sydney 2017
Telcos are flocking to OpenStack for it's NFV functionalities, ability to be deployed in modular POD architectures with federated deployments across datacenters, central offices & edge services. But to deploy something that can meet the business needs in the Telco world, there's alot of crucial design decisions that must be made. What's the right hardware? Which SDN and/or SDS option meets our performance needs? What topology delivers the right amount of High Availability? How can we make all of these decisions, yet make it cost effective? Come join us & hear our lessons learned deploying OpenStack at large Telcos. We'll talk about topics such as L2/L3 stretching between OpenStack PODs, Service Provider integrations, identity federation, deployment automation, service function chaining & federated overlay networks. We'll share pitfalls & issues encountered along the way but more importantly, we'll tell you how we were able to overcome these hurdles so you can too!
Automated NFV deployment and management with TripleO - OpenStack Summit Sydney 2017
In this session, learn about the critical aspect of NFV enablement in OpenStack deployments: how to automatically & continuously manage the platform & network optimizations for NFV. Focusing on creating a seamless user experience, learn about the latest features and efforts underway to leverage composable roles to simplify the life cycle (enable, disable, & change) management and provide greater control to the VNF user. We’ll also share valuable insight into the mistakes made and lessons learned that helped us along the way.
No Longer Considered an Epic Spell of Transformation- IN-PLACE UPGRADE! - OpenStack Summit Boston 2017
Welcome to the magical school of Triple-O where impossibilities are not only possible, but a reality! Upgrade from Mitaka to Newton? Sure, we can do that and YOU can too! Let us show you the value of TripleO Magic! In place upgrades is only one item in our bag of tricks (that can be YOUR bag of tricks too!) Come join us while we show you how to conjure a rock solid OpenStack environment that has continuous validations running to ensure stability throughout the entire lifecycle of that cloud even THROUGH your in-place upgrade! This isn't a cookie cutter deployment either! Customizations abound! The reference architecture not working for you? Use your magic wand to create custom composable roles and deploy YOUR cloud YOUR way!
Feeling a Bit Deprecated? We Are Too. Let's Work Together to Embrace the OpenStack Unified CLI!- OpenStack Summit Barcelona 2016
Does anyone else recall learning the CLI of the OpenStack projects and scratching their head in bewilderment as to the variance in naming, structure, and arguments for the various projects? And then the output you received from executing the command was different depending on which project you were running against?
Enter the OpenStack Unified CLI to level set these commands. The Unified CLI not only attempts to create a consistent feel between the OpenStack projects, it also is a pluggable architecture to dynamically provided commands, depending on which OpenStack components you had installed. For instance, you don't get OpenStack bare metal unless you have Ironic installed.
In this session we will deep dive into the OpenStack Unified CLI. We’ll show how it works behind the scenes, and we’ll discuss its available plugins. We’ll also provide some troubleshooting guidance for those times when things don't work as expected.