chapter-9-log-and-monitor-5.md

Table of Contents

Chapter 9: Log and Monitor (5%)

• Monitor All Cluster Component
• Monitor Applications
• Cluster Component Logs
• Manage Application Logs

Log and Monitor

Monitor All Cluster Component

The Metrix server allows to collect CPU and memory data from the nodes and pods in the cluster. Install metrix server

-- clone metrix server repository
$ git clone -b release-0.3 --single-branch https://github.com/kubernetes-sigs/metrics-server.git

-- install metrix server
$ kubectl apply -f metrics-server/deploy/1.8+/

-- get metrix server information
$ kubectl get --raw /apis/metrics.k8s.io

Command References

 -- view CPU and memory for the nodes
 $ kubectl top node 
 
 -- view CPU and memory for the pods
 $ kubectl top pods
 
 -- view CPU and memory for the pods in all namespaces and kube-system
 $ kubectl top pods --all-namespaces
 $ kubectl top pods -n kube-system
 
 -- view CPU and memory for the pods with selectors
 $ kubectl top pods -l run=pod-with-defaults 
 
 -- view CPU and memory for the pods with pod_name
 $ kubectl top pod pod-with-defaults 
 
 -- view CPU and memory for the pod's container
 $ kubectl top pods group-context --containers 
 

References and Further Study

• https://kubernetes.io/docs/concepts/cluster-administration/cluster-administration-overview/#securing-a-cluster

Monitor Applications

Liveness and Readiness Probes

Liveness and readiness probes can be used to automatically restart containers if they are failing or automatically remove from a service point.

Command References

 -- get service endpoint
 $ kubectl get ep

References and Further Study

Cluster Component Logs

The log directory for containers is in /var/log/containers, which can potentially can consume all nodes disk spaces. To manage this use sidecar technique with a logging agent.

Command References

 -- view the different type of logs 
 $ kubectl exec counter  -- ls /var/log

 -- view logs of a container in a pod
 $ kubectl logs counter conunt-log-1
 $ kubectl logs counter conunt-log-2

References and Further Study

Manage Application Logs

Command References

 -- view the logs of nginx container
 $ kubectl logs nginx 

 -- view logs of a container in a pod
 $ kubectl logs counter -c conunt-log-1

 -- view logs of all contianers
 $ kubectl logs counter --all-containers=True

 -- view logs of a container in a pod labeled nginx
 $ kubectl logs -p -c nginx nginx

 -- view logs of previously terminated container on pod nginx
 $ kubectl logs counter conunt-log-1

 -- tails the log for a pod
 $ kubectl logs --tail=20 nginx

 -- view the last 1 hour logs
 $ kubectl logs --since=1h nginx

 -- view the contianers logs from a deployment 
 $ kubectl logs deployment/nginx -c nginx

 -- direct logs to a file
 $ kubectl logs counter -c count-log-1 > counter.log

References and Further Study

Table of Contents
Prev Chapter: Chapter 8: Security (12%)
Next Chapter: Chapter 10: Troubleshooting (10%)