Shine a Light on K8s Blindspots With Komodor and Anodot

Transcript

1. Komodor <> Epsagon | May 2021
   Shining a Light on
   K8s Blindspots
   Nir Shtein, Software Engineer @ Komodor
   

   View Slide

2. Investors | January 2021
   Investors | January 2021
   I’m Nir Shtein
   ● Software Engineer at Komodor
   ● The main contributor to ValidKube.com
   ● Open-Source advocate
   ● Kubernetes fan :)
   Hi, nice to meet you!
   

   View Slide

3. Cloud native | March 2021
   What challenges does Kubernetes present?
   Scattered data
   Distributed systems make it harder to
   understand the effect of a single change
   Lack of K8s knowledge
   The knowledge & expertise around K8s
   is often held by only a few (i.e DevOps).
   Blind spots
   Changes are unaudited
   or hidden
   

   View Slide

4. Cloud native | March 2021
   "Despite 6 years of progress, Kubernetes is still incredibly complex," said Drew Bradstock, product lead for
   Google Kubernetes Engine (GKE). "What we've seen in the past year or so is a lot of enterprises are embracing
   Kubernetes, but then they run headlong into the difficulty."
   And the common thread is: Complexity
   

   View Slide

5. Cloud native | March 2021
   Kubernetes Is like a car…
   UI INFRA
   

   View Slide

6. Cloud native | March 2021
   Or like an iceberg…
   

   View Slide

7. Cloud native | March 2021
   So how can we simplify
   K8s troubleshooting?
   

   View Slide

8. Investors | January 2021
   Investors | January 2021
   Make sure to include important metadata:
   ● Labels and annotations
   ● Environment variables
   ● Secrets
   ● Config maps that point to the proper
   objects and volumes
   ● Configuring liveness probes
   ● Readiness probes
   Best Practice #1:
   Maintain a Good YAML Hygiene
   

   View Slide

9. Investors | January 2021
   Investors | January 2021
   Tag and label your logs properly, by including the:
   ● Proper service name (not the pod names!)
   ● Version
   ● Cluster environment information
   ● Business-specific data
   Best Practice #2:
   Logging - Specifically for K8s
   

   View Slide

10. Investors | January 2021
    Investors | January 2021
    There are several ways to separate your K8s environments:
    Option 1: Create an environment for each stage of the
    development process (development, QA, staging, production)
    Best Practice #3:
    Separate/Segregate Environments
    

    View Slide

11. Investors | January 2021
    Investors | January 2021
    Option 2: Create an environment according to namespaces
    (special K8s resource)
    Best Practice #3:
    Separate/Segregate Environments
    

    View Slide

12. Investors | January 2021
    Investors | January 2021
    Best Practice #4:
    Invest in Proper Monitoring
    Open Source
    Monitoring Solutions
    Step 1: Choose the right monitoring solution for your needs.
    Commercial
    Monitoring Solutions
    

    View Slide

13. Investors | January 2021
    Investors | January 2021
    Best Practice #4:
    Invest in Proper Monitoring (Cont’d)
    Step 2: Start monitoring the following metrics:
    ● Resources: CPU / Memory Usage
    ● Container Status: Up / Down / Errors / Probe Data / Restart
    count
    ● Application Metrics: Application Performance Metrics - APMs
    Make sure to monitor these metrics in an automated way by setting up
    proper monitors and alerts
    

    View Slide

14. Cloud native | March 2021
    Illuminate the Kubernetes darkness
    

    View Slide

15. Automate K8s Troubleshooting with Komodor
    

    View Slide

16. Investors | January 2021
    Investors | January 2021
    What dark areas of Kubernetes
    can Komodor illuminate?
    ● Multi-cluster visibility
    ● Historical view
    ● Node issues
    ● Deleted pods
    

    View Slide