clovis
@clovistb
Followers
1K
Following
5K
Media
116
Statuses
2K
Chasing Kubernetes wisdom • DevOps Engineer • laC addict • Building platforms
Houston, TX
Joined March 2011
I have two EC2 instances. EC2-A → client EC2-B (10.0.2.15)→ server running an app on port 8080 From EC2-A: curl http://10.0.2.15:8080 → works ✅ ping 10.0.2.15 → fails ❌ Security group is open for TCP 8080. Why does ping fail while curl
28
15
197
Tip about Amazon CloudFront and AWS Global Accelerator - Use CloudFront when the main problem is distance between users and your content. CloudFront caches files (images, HTML, API responses, videos) at edge locations around the world so users receive them from the closest
CloudFront vs Global Accelerator (basic explanation) You deploy a resource (EC2, S3, ALB, etc.) in us-east-1 and you want to serve users in Europe. Instead of letting European users connect directly to us-east-1 over the internet, AWS provides two solutions. 1. CloudFront
0
2
23
CloudFront vs Global Accelerator (basic explanation) You deploy a resource (EC2, S3, ALB, etc.) in us-east-1 and you want to serve users in Europe. Instead of letting European users connect directly to us-east-1 over the internet, AWS provides two solutions. 1. CloudFront
4
8
49
If you’re still wondering about TCP vs ICMP, read the post below and the comments. The community already explained everything.
I have two EC2 instances. EC2-A → client EC2-B (10.0.2.15)→ server running an app on port 8080 From EC2-A: curl http://10.0.2.15:8080 → works ✅ ping 10.0.2.15 → fails ❌ Security group is open for TCP 8080. Why does ping fail while curl
1
2
19
DevOps is easy. You just need to master a few tools 😄 Git npm Jira Snyk Chef Helm Asana Trivy Falco Maven Nomad Slack Zabbix Docker Kibana Trello GitHub GitLab Puppet Pulumi Nagios Splunk Jaeger Harbor Tekton Podman Bamboo Argo CD Flux CD Ansible Rancher Jenkins Grafana
29
49
257
Imagine 100 users open a website at the same time and ask a simple question in their browser: “How many months are in a year?” The system takes 250 milliseconds to reply “12 months.” Out of the 100 users, 98 receive the correct answer, while 2 users do not receive a response.
4
4
19
7 Steps to Reduce Latency in Any Web System. Reducing latency in a system means reducing the time between when a user sends a request and when the system returns a response 1️⃣Reduce network distance ◦Deploy services closer to users (multi-region architecture) ◦Use a CDN to
0
4
26
If you’re still struggling to understand the four pillars of monitoring - latency - traffic - errors - saturation, The example below may help👇
Imagine 100 users open a website at the same time and ask a simple question in their browser: “How many months are in a year?” The system takes 250 milliseconds to reply “12 months.” Out of the 100 users, 98 receive the correct answer, while 2 users do not receive a response.
1
2
13
My opinion I would proceed by elimination. ❌Cloud Engineer: Strong in infrastructure, but not always polyvalent.A DevOps or Platform Engineer can cover cloud tasks. ❌SRE: Excellent in reliability, monitoring, and incident response. But may not replace a Cloud Engineer or
5
1
25
Imagine 100 users open a website at the same time and ask a simple question in their browser: “How many months are in a year?” The system takes 250 milliseconds to reply “12 months.” Out of the 100 users, 98 receive the correct answer, while 2 users do not receive a response.
4
4
19
I fixed incidents and then turned them into LLM tasks to evaluate how well AI agents can fix the same issues.
What is the one thing that you did with AI that you are proud of? I was literally asked this in an Interview.
0
1
6
Your system is down and you’re looking for a temporary fix. First question: Did we deploy or change anything recently? If the answer is no, don’t think about rollback. Rollback is only useful when the outage is caused by a recent change. It should not be your first option.
8
2
26
The difference between a junior and an experienced engineer. The experienced engineer performs well in the first few months, then starts believing he already knows everything. The junior engineer costs you the first few months, but keeps pushing because he feels he still has
3
1
22
Never underestimate a post you create. It can go farther than you ever imagine. Even if there are no royalties on posts, the idea is still yours. The work is still yours.
My post got copied on LinkedIn. Looks like my content finally got its promotion to the corporate recycling center. 😄
1
0
7
Production rule If your system is down, check in this order: 1 - ALB 2 - ECS Tasks 3 - Database Because failures usually top → bottom.
1
1
7
When running an ECS + Fargate architecture, monitoring should happen at three levels. 1 - ALB (Application Load Balancer) This is the entry point of your system. Monitor: - Request Count → traffic level - Target Response Time → latency - HTTP 5xx errors →
3
6
39
What is the difference between horizontal and vertical scaling? If you want to stay above the average candidate, don’t repeat the textbook answer. Interviewers are testing if you think at the system level, not just the resource level. Think scaling at: - Server level -
2
0
18