Kubernetes

This article was published originally at https://kubevirt.io/2020/Live-migration.html Introduction This blog post will be explaining on KubeVirt’s ability to perform live migration of virtual machines. Live Migration is a process during which a running Virtual Machine Instance moves to another compute node while the guest workload continues to run and remain accessible. The concept of live migration is already well-known among virtualization platforms and enables administrators to keep user workloads running while the servers can be moved to maintenance for any reason that you might think of like: ...

This article was published originally at https://metal3.io/blog/2019/09/11/Baremetal-operator.html Introduction The baremetal operator, documented at https://github.com/metal3-io/baremetal-operator/blob/master/docs/api.md, it’s the Operator in charge of definitions of physical hosts, containing information about how to reach the Out of Band management controller, URL with the desired image to provision, plus other properties related with hosts being used for provisioning instances. Quoting from the project: The Bare Metal Operator implements a Kubernetes API for managing bare metal hosts. It maintains an inventory of available hosts as instances of the BareMetalHost Custom Resource Definition. The Bare Metal Operator knows how to: Inspect the host’s hardware details and report them on the corresponding BareMetalHost. This includes information about CPUs, RAM, disks, NICs, and more. Provision hosts with a desired image Clean a host’s disk contents before or after provisioning. ...

After some time checking the scenarios at https://learn.openshift.com, I decided to give it a try. With the help of Mario Vázquez, author of Getting Started with Kubefed, I did create two scenarios: How to use Citellus on Citellus: Troubleshooting automation KubeVirt on a ‘browser-based’ approach for MiniKube setup for validating KubeVirt: Kubernetes with VM Virtualization (versus the regular containers). You can check how them can be created by looking at their code at: killercoda Scenarios or the ‘playable’ version at https://killercoda.com/iranzo/. ...

Postgres across clusters For Postgres to work across clusters we do need to have the data being synchronized. With some other databases we do have some master-master approach, but usually have very strict requirements on latency, bandwidth, etc that we cannot solve with On-Premise + external cloud providers. If the replication is based on the storage level instead, then you face that database servers don’t deal well if the data changes underneath it, so it leads to data corruption, on top of the storage-level issues/requirements as well on bandwidth, latency, etc. ...

Why this article? For Federation of OpenShift/Kubernetes clusters we want not only to demo some applications, but build a solution that covers all that will be needed in a real-world deployment. Colleagues in the Solutions Engineering team have been working on demonstrating an application running on different clusters over a Federated Control Plane, allowing it to ‘roam’ between clusters with shared data access as a way to demonstrate how an application can ‘scale’ from on-premise deployment to external clouds in order to satisfy business requirements for peak demands. ...