In the world of SRE, we talk a lot about observability, such as metrics, logs, and traces. In most cases, these three pillars of observability are related to infrastructure metrics or business metrics. For example, the current CPU usage of a Kubernetes pod, the number of videos played by online users per minute, etc. Besides this, there is another important area that is being explored nowadays: FinOps.
Posts under category CloudOps
1. Check the partitions information of the target disk.
sudo fdisk -l
sudo parted -l
This blog presents how to deploy a self-hosted Kubernetes cluster using kubeadm.
Note that usually we use kubeadm to bootstrap a K8s cluster that has several nodes. It is also easy to scale up the created cluster using kubeadm later. However, if you only need to have a learning environment of K8s, it might be easier to use some tools like kind, and minikube: https://kubernetes.io/docs/tasks/tools/
By default, there is only one network created in an OpenStack project. But for some scenarios, you might need to create several instances and organize them in an internal network. This internal network could be connected to the public Internet through a router (more discussions in this issue). This blog demonstrates how to create such an internal network in an OpenStack project. A corresponding tutorial video could be found at https://www.youtube.com/watch?v=3RmUpWA7U1Q
Sometimes when we use a virtual machine on some cloud platforms such as Azure, we might face the following scenario. The hard disk is running out of space. Thus we expand the size of that hard disk from the cloud platform's dashboard. After logging in to the system again, we find that the size of the disk is successfully expanded while the partition remains the same size.