Posts from @houfu.lee..
Organizations increased GPU instance spending by 40%, primarily using the cost-effective G4dn type. Arm-based instances, like the T4g with Graviton2, see a rise in usage for their energy efficiency and performance. Container use has grown to 35% of EC2 spend, though 83% of container costs stem from .. read more
The post explains how to add authentication to APIs using Amazon Cognito in AWS. It covers concepts like User Pool, User Pool Domain, Resource Server, and User Pool Client. The setup involves creating an Auth Stack containing the User Pool and User Pool Domain resources, and then sharing the data to.. read more
Apache Superset is an open-source dashboard solution that allows for easy sharing and embedding of BI charts. It features a variety of chart types, a powerful query builder, and support for custom connectors. Many big companies use Superset, and installation on Kubernetes is made simple with just tw.. read more
Global spending on observability is over 2.4 billion USD in 2024 and expected to reach 4.1 billion by 2028. The growth of observability costs is driven by factors like microservices, ephemeral servers, and SRE/Chaos Engineering practices. To reduce observability costs, companies can consider impleme.. read more
The dotdc/grafana-dashboards-kubernetes project has seen significant growth over the past two years, with increased traffic and new contributors, including users from companies like Rakuten, Orange, Swisscom, and Nokia. The project now features proper releases generated using semantic versioning, cl.. read more
This post serves as a quick reference for changes in each Python version, helping users leverage new features or maintain code compatibility. It covers syntax changes, standard library modifications, and useful tools for upgrading code bases. Each Python version is detailed with major changes, end-o.. read more
DevOps, introduced around 2007, aimed to merge development and operations teams to streamline the release process. While this promised increased deployment speed and feedback, the reality revealed challenges in maintaining consistency and ownership of issues. The adoption of containers provided a si.. read more
This blog post explains how Round Robin Escalations can improve on-call scheduling by distributing the workload amongst a team of responders. It highlights the benefits of this approach such as fairer workload distribution, faster response times, and reduced stress for on-call staff. The blog also details who can benefit from Round Robin Escalations, including support teams and IT operations teams, and concludes by explaining how this system works.
This blog post argues that collaboration between developers and SREs is essential for building reliable software. The blog post outlines five ways that developers can improve SRE observability:
Embrace the 12-Factor App Methodology: This methodology creates applications that are easier to deploy and monitor.
Share Performance Testing Data: This data helps SREs understand how the application should function under pressure.
Maintain Clear and Concise Documentation: Clear documentation empowers SREs to resolve issues faster.
Leverage AIOps for System Administration: AIOps automates tasks and improves IT operations.
Increase System Observability Through Code: Expose relevant metrics within the code to provide SREs with real-time insights.
Have a nice weekend! But first, look at this article that provides a step-by-step guide for updating and upgrading the RELIANOID Load Balancer Cluster with zero downtime. Depending on the services configured, you may experience little to no disruption. However, user sessions might be impacted based ..
