Alexis Dufour

As part of the ITOPS team, our main mission is the management of all FLOA infrastructure on Azure cloud.

• For all Azure resources, we have industrialized their deployment through the use of Terraform and Terragrunt

• Implementation of a "Landing Zone" to properly separate each project and access based on needs

• Implementation of a multi-zone and multi-region resilient infrastructure (via Hub & Spoke)

• Implementation of "Disaster Recovery" using Azure tools such as Azure Site Recovery

• Management of Kubernetes clusters (AKS) and use of the Gitops methodology (ArgoCD) for application deployment

• Implementation of monitoring stacks such as Prometheus/Grafana/Thanos

• Implementation of CICD via GitlabCI, for the automation of all our deployments

• Use of Ansible and AWX for configuring VMs on Azure cloud

• Automatic management of updates for our tools and infrastructure

• Implementation of FinOps technical strategies on Azure aimed at significantly reducing infrastructure costs:

• Automation of startup/shutdown of non-production environments during nights and weekends

• Use of Spot instances for non-critical environments, generating up to 70% savings

• Analysis of actual consumption to size purchases of reserved instances and Savings Plans, allowing an average reduction of 50% in monthly costs related to Azure resources

As part of the IT4Container team, our goal is to provide turnkey Kubernetes clusters for OUISncf employees (On Premise, Cloud(AWS)).

As well as giving them maximum autonomy, from cluster creation with a single click to the creation of a namespace specific to each application.

• Implementation of industrialization to facilitate the installation, update, or addition of "workers" to a Kubernetes cluster with Kubeadm. Use of Ansible to automate tasks end-to-end (Currently managing about ten Kubernetes clusters).

As part of the Cloud migration, creation of EKS (AWS) clusters using "Eksctl". End-to-end automation with Ansible. We currently maintain about thirty clusters.

Implementation of AWX to make Ansible "API-able" and give users more autonomy.

• In the OUISncf context, development of "controllers" and "admission controllers" in Go, which aim to inject "sidecars" (for log reading) or "labels/annotations" for billing purposes on Kubernetes resources.

• Implementation of the GitOps methodology with the FluxCD tool for all tools installed on a cluster. Use of Helm and Kustomize for templating.

• Implementation of a security policy via NetworkPolicy and RBAC specific to each namespace. Use of OPA Gatekeeper to block deployments that do not comply with security rules.

• Implementation of a monitoring stack (Grafana/Prometheus/AlertManager and Datadog) on all Kubernetes clusters. Use of "ServiceMonitor" and creation of "PrometheusRule" to forward alerts to AlertManager.

Use of Thanos for long-term storage.

• Use of Terraform for the initialization of AWS resources (except EKS).

Platform connecting professionals during their business trips to be hosted for free (www.tepee.pro)

• Development in ReactJs and Ruby On Rails

• Cloud platform management (Google Cloud Platform, Kubernetes)