How To Fix OOMKilled

Good day. It's Monday, Aug. 12, and in this issue, we're covering:

• How To Fix OOMKilled

• HashiCorp 2024 State of Cloud Strategy Report

• Google Cloud launches BigQuery continuous queries

• Adidas Reduced Kubernetes Cluster Costs by 50%

• Bash Script to Install Multiple Helm Charts in One Go

• Terraform Ephemeral Values Explained

You share. We listen. As always, send us feedback at [email protected]

Use Case

How To Fix OOMKilled

OOMKilled occurs in Kubernetes when a container exceeds its memory limit or tries to access unavailable resources on a node, flagged by exit code 137.

Typical OOMKilled looks like

NAME                READY     STATUS        RESTARTS     AGE
web-app-pod-1       0/1       OOMKilled     0            4m7s

Learn more about OOMKilled visually here:

How to Fix OOMKilled Kubernetes Error (Exit Code 137)

1. Identify OOMKilled Event: Run kubectl get pods and check if the pod status shows OOMKilled.

2. Gather Pod Details: Use kubectl describe pod [pod-name] and review the Events section for the OOMKilled reason.

Check the Events section of the describe pod, and look for the following message:

State:          Running
       Started:      Mon, 11 Aug 2024 19:15:00 +0200
       Last State:   Terminated
       Reason:       OOMKilled
       Exit Code:    137
       ...

3. Analyze Memory Usage: Check memory usage patterns to identify if the limit was exceeded due to a spike or consistent high usage.

4. Adjust Memory Settings: Increase memory limits in pod specs if necessary, or debug and fix any memory leaks in the application.

5. Prevent Overcommitment: Ensure memory requests do not exceed node capacity by adjusting pod resource requests and limits.

Point worth noting:

To check the QoS class of a pod, run this command:

 kubectl get pod -o jsonpath='{.status.qosClass}' 

To inspect the oom_score of a pod:

1. Run kubectl exec -it /bin/bash

2. To see the  oom_score, run cat/proc//oom_score

3. To see the  oom_score_adj, run cat/proc//oom_score_adj

The pod with the lowest oom_score is the first to be terminated when the node experiences memory exhaustion.

Tool Of The Day

An open-source antivirus engine for detecting trojans, viruses, malware & other malicious threats.

Try it out !

Trends & Updates

HashiCorp 2024 State of Cloud Strategy Report

The report reveals only 8% of organizations are highly cloud-mature, gaining stronger security and faster development. Key findings include 91% reporting cloud waste, 64% facing a skills shortage, and 79% using or planning multi-cloud deployments.

Google Cloud launches BigQuery continuous queries

This new feature allows instant data processing, integrates with Google’s AI tools for real-time machine learning, and simplifies event-driven architectures, all within BigQuery.

Resources & Tutorials

Adidas Reduced Kubernetes Cluster Costs by 50%

By automating resource scaling with Karpenter, enhancing efficiency with Kyverno-driven VPAs, and strategically downsizing during non-peak hours, all while tackling challenges like node underutilization and balancing application performance.

Bash Script to Install Multiple Helm Charts in One Go

This article introduces "helmister," a lightweight bash script that automates the installation and uninstallation of multiple Helm charts, ideal for secure, air-gapped environments where simplicity and accessibility are key.

Terraform Ephemeral Values Explained

Explore how ephemeral values in Terraform can minimize secret sprawl and enhance security by limiting the persistence of sensitive data, offering a fresh perspective on managing secrets and other transient resources effectively.

Did someone forward this email to you? Sign up here