Most of my work starts when something in production looks wrong: traffic shifts, alerts get noisy, a route breaks, cache behavior changes, or a service just stops acting normal. I spend my time tracing those problems across CDN, DNS, network and cloud layers, getting the service stable again, and improving the monitoring or automation around whatever caused the pain.
During an incident, I care about finding what is actually failing, not just what happened to alert first. That usually means checking traffic, service health, logs, headers, cache behavior, DNS, origins and network paths, then pulling in the right team when the problem crosses a boundary.
I use Prometheus, Grafana and ELK every day, but the tool matters less than whether it helps answer the right question. I spend a fair amount of time tuning noisy alerts, building dashboards around real operational needs, and trying to shorten the gap between "something is wrong" and knowing why.
If a repetitive operational task can be made safer or faster, I usually try to automate it. My day-to-day toolbox includes Linux, Docker, HAProxy, OpenStack, Ansible and Terraform, plus scripting and small internal tools when the existing ones are not enough.
Sep 2024 — Present. I moved into the NOC team after about two years on the CDN product side. These days I'm much closer to production: watching service health, handling incidents, investigating DDoS and unusual traffic, and debugging issues across CDN, DNS, storage, network and cloud services. One thing I care about a lot is making the next incident easier to understand than the last one — better alerts, clearer runbooks, better handoffs and less guesswork.
Aug 2022 — Sep 2024. Before NOC, I worked on the CDN product team. That role gave me a useful view of how infrastructure decisions show up on the customer side. I worked on onboarding, usage analysis, abuse and revenue-protection cases, WAF improvements, release checklists and integrations such as WordPress and WHMCS. I also spent time researching API Gateway, Edge Computing and Zero Trust ideas.
Aug 2021 — Apr 2023. MACSA was where I got to build and own more of the stack myself. I designed a multi-branch L2/L3 network, set up monitoring and centralized logs, deployed VoIP, storage and backups, documented failover and incident procedures, and kept the environment running. After joining ArvanCloud, I continued maintaining that infrastructure remotely for several months.