About
~ cat about.md
I don't manage servers by hand: I build the chain that manages them for me.
14 years in infrastructure, including 6 designing and running Kubernetes clusters. Private cloud, no dependency on a hyperscaler.
- 14years in infrastructure
- 6years of Kubernetes
- 300+application instances run
- 21clusters on private cloud
My job: build the platform and the tooling that make production reproducible, so development teams can ship without depending on me.
I design and build from scratch the Kubernetes platform of a software company, and the whole automation chain that deploys it. Every change goes through a merge request; the tooling deploys it, never a manual action on the cluster.
The infrastructure I run is remote, deployments go through code and GitOps, and the team can work asynchronously. My work day is the same in an office or at home: I work full remote from Marseille, with occasional travel on site.
Technical skills
~ kubectl get skills
| NAME | STACK | SINCE |
|---|---|---|
| orchestration | Kubernetes · Talos · Cilium eBPF · Helm · Docker | 2021 |
| infrastructure-as-code | Ansible · Terraform · Terragrunt · SaltStack | 2019 |
| gitops | GitLab CI/CD · ArgoCD · Jenkins · Nexus | 2021 |
| observability | Prometheus · Thanos · Grafana · Loki · Graylog | 2021 |
| storage-and-data | Ceph · S3 · PostgreSQL · MongoDB · OpenStack · VMware | 2019 |
| security | Network segmentation · External secrets · Least privilege · Traceability | 2014 |
| tooling | Python · Go · Bash | 2019 |
Work experience
~ git log --oneline --author="cyril" -- experiences/
Click a line to open the details.
-
~ git show a7f3c21
Kubernetes platform of a healthcare software company, on private cloud. The platform is run end to end, from the virtualization layer to application delivery, with no manual action on the clusters. - Clusters on an immutable distribution with an eBPF CNI, provisioned with Infrastructure as Code, with no interactive access to the nodes - Ansible automation: reusable roles, inventories per environment, runs driven by a scheduler - Continuous deployment with GitLab CI/CD and ArgoCD: every change goes through a merge request, never through a manual action on the cluster - Helm packaging: versioned internal charts, automated dependency updates, release pipeline - Controlled supply chain: images from an internal registry, pinned versions, vulnerability tracking - Secrets lifecycle moved to an external vault - Multi-cluster observability: metrics, logs, alerting and long-term retention on object storage - Backup and restore of clusters and data - Migration of legacy applications to the target platform, with the same behavior - Operations and diagnostic tooling written in Python and Bash
~
-
~ git show 4b19e80
Service continuity on the cloud infrastructure of the PC-SCOL project. Deployment of the Pegase software on more than 300 Kubernetes instances across several clusters, at an OpenStack provider. - Development of the application tooling, written as Infrastructure as Code - Advanced support on the cloud infrastructure and CI/CD - Network administration - Maintenance and evolution of internal tools Stack: Kubernetes, Docker | Helm, Terraform, Terragrunt | Python, Go, Bash | GitLab, Nexus | Jenkins, ArgoCD | Grafana, Prometheus, Loki, Thanos | PostgreSQL, MongoDB | S3 Training: Docker and Kubernetes (Enix), ArgoCD and Golang (Udemy)
~
-
~ git show c5d2a13
Deployment of Linux server clusters with Infrastructure as Code, with Python development. International assignments for various clients, mainly telecom operators, with responsibility for service continuity. - Projects and development of production tools: Python, GitLab, SaltStack, Ansible, Debian, Jenkins, Vagrant, Libvirt, Inspec - Linux server operations: Debian, SaltStack, GitLab, MariaDB, Redis, Icinga, Grafana, Nagstamon, Nginx, Netfilter, Dell servers, Cisco switches - Constant attention to security and network management Training: Kubernetes (M2i), Terraform self-study, Spanish B1 (Arca delle Lingue)
~
-
~ git show 9e0b7f4
In charge of the network, system and telephony infrastructure of three sites in La Ciotat and one site in Besançon. - Network: Cisco Nexus, HP Procurve, Ubiquiti Unifi, Linux routing and firewall, Iptables, OpenVPN, Bind, IPAM - System: VMware ESXi, Nimble Storage, Synology, Debian, Ubuntu Server, Windows 2012R2, AD, DNS, GPO, MDT/WDS, SQL Server - Operations: Veeam, Nagios, Observium, Graylog, Oxidized, GLPI, Zimbra, XIVO, Office 365 Training: English with Wall Street English (2018), Nimble Storage certification and SQL Server administration (2017), ZFS and Cisco Nexus (2016), Linux security (2015) Before 2014: CPAM Essonne, Initem, Pôle Sud.
~
Contact
~ ./contact.sh
Marseille, France · full remote