Introduction According to the Kubernetes documents, when a container in a pod consumes more memory than its limit, the pod is killed. A Pod killed by OOM (OOMKilled Pod) will be restarted if the r...

Introduction When deploying application on Kubernetes cluster, people may use NodePort service to make their application accessible. Encountering the error message failed to allocate a nodePort: r...

Introduction Depending on the use case, sometimes people would use hostPath to mount a file or directory from the host node’s filesystem into their Pods.1 An interesting observation is that when a...

Introduction Sometimes when containers running from the same image show different starting time in different Kubernetes cluster, we would normally first compare the different resources in the clus...

Introduction Sometimes you may wonder which process is responsible for sending the DNS request on your servers. While capturing DNS packets with tcpdump is possible, identifying the process ID(PID...

Introduction Amazon ECS has being supporting Amazon Elastic File System(EFS) over the years.1 AWS provides resources including tutorials and several Knowledge Center articles and video2 to guide c...

Introduction / Summary Occasionally, customers may experience their EC2 instances failing the instance status check1 unexpectedly. However, these instances typically return to a healthy state afte...

Introduction Over the past year, my role and responsibilities as an AWS Cloud Support Engineer has significantly expanded. Apart from handling daily cases, I’ve been focused on enhancing AWS Knowl...

Introduction This article will go through what I learned from How Amazon ECS manages CPU and memory resources and few things I add about the CPU and memory management of Amazon ECS. Two General R...

Introduction / Summary This article will go though a brief explanation that why storage_stats and blkio_stats have different output from ECS metadata endpoint. Scenario Some people will question...