Margaret Hoagland’s Post

Learn how extensions can help you independently scale your #OpenSearch workloads & right-size your resources, without impacting cluster availability. New on the blog: http://spr.ly/6044PnBx4

Why should we care about database monitoring? 🙂 🤔 The following blog post is about #PMM, and how we can take advantage of this Open Source tool to monitor our databases and prevent problems on time. With PMM, we will see how to monitor: - Long Query Response Times - High CPU Utilization - High Memory usage - Disk I/O - User connections - Data growth And much more 💃🏽 Check out my next blog post on the Percona community 🌱 Do you wanna test PMM? Are you using it already? Let me know how I can help 😊 #database #percona #opensource #monitoring

Best Caching and Performance Optimization Techniques

💪 This blog offers a comprehensive guide to scaling Kong deployments with and without databases. Get your hands on the tips and tricks now! #KongPin #Scaling #Deployments #TipsandTricks

Post #3 Title: Troubleshooting Kubernetes: Solving Pod Scheduling Issues Content: Today, I encountered an interesting challenge in one of our Kubernetes clusters. A pod was stuck in a pending state, and upon further investigation using kubectl describe, I discovered the error message "5 Insufficient cpu. preemption: 0/6 nodes are available: 1 Preemption is not helpful for scheduling, 5 1.." From the error message its pretty much understandable some issue with the nodes CPU. But when I checked, nodes appearing fine in terms of CPU and memory usage ,but still the pod scheduling issue persisted. Solution Journey: Fortunately, that cluster was integrated with Dynatrace for monitoring. I deep dived into the metrics and discovered a whopping 130 workloads spread across 25 namespaces. However, no quotas were set at the namespace level. Next, I scrutinized the CPU usage, requests, and limits for all workloads. A significant difference between actual CPU usage and requested resources caught my attention. Reviewing the past 7 days' CPU usage data for these workloads, I recalibrated resource requests and limits accordingly. This adjustment effectively freed up CPU and memory resources within the cluster. And Pods are able to schedule properly and come into the running state. Key Takeaway: This experience reinforced a valuable lesson: setting Kubernetes resource limits requires precision. Incorrect values can impact both cluster nodes and application performance. Remember, finding the right balance is crucial. Here's a helpful tip: leverage open-source tools like Goldilocks (https://lnkd.in/dA7jgznh) to streamline this process. Let's continue to optimize and fine-tune our Kubernetes environments for peak efficiency! 💡🚀 #kubernetes #troubleshooting #devopsengineer

You also don’t need much experience with Kubernetes administration to install nginx using our managed service. This tutorial is based on GUI (graphical user interface) tools, so no manual manipulations are required in the command line.

There are paths to #mainframe optimization, but it’s not an all-or-nothing proposition. Learn five #mainframemodernization best practices in our blog post. https://okt.to/wUX2lJ

#StarWind_handy #Backup basics: Full, Incremental or Differential? Learn about the types in the new article by Diana Abo-Harmush for #StarWind and choose the right strategy for your #datasecurity. https://lnkd.in/eDucp53z