Over the years I founded and ran several IT businesses — winning clients and delivering enterprise infrastructure for organizations of 200–1,000+ users. I built and led the technical teams behind that work: hiring, interviewing, and mentoring engineers, directing sysadmins at client sites across several countries, and reporting directly to C-level — all while staying hands-on in the systems myself. Across ~30 years I've worked at a strong level as both a Windows systems administrator and a Linux engineer — server estates, networks, virtualization (VMware ESXi) — and moved into containerization, platform engineering, Kubernetes, cloud, and DevSecOps. CKA and CKS certified, AWS Solutions Architect – Professional (in progress). Based in Tel Aviv.
Infrastructure & Security
3-node kubeadm cluster · Calico CNI · containerd · k8s v1.35
$ kubectl get pods -n production web-7d9f8b-xk2p 1/1 Running api-6c8df9-m3nw 1/1 Running HPA min:2 max:10 cpu:42% PDB minAvailable: 1
Production-grade manifests — default-deny NetworkPolicy, least-privilege RBAC, zero-downtime RollingUpdate, HPA with scale-down stabilization, PDB, and a full Helm chart.
GitHub$ kubectl get networkpolicies -A default default-deny-all default allow-dns-egress production allow-ingress-nginx PSA: restricted · audit: on
Zero-trust network segmentation, least-privilege RBAC, Pod Security Admission, API server audit policy, and custom seccomp profiles — all on a physical kubeadm cluster.
GitHub$ kubectl get prometheusrule -n monitoring homelab-custom-alerts Active NodeMemoryPressureHigh EtcdHighCommitDuration PodCrashLooping
kube-prometheus-stack tuned for a RAM-constrained homelab — custom PrometheusRule CRDs for node, etcd, and pod-lifecycle alerts. Alertmanager routing with push notifications.
GitHub$ argocd app list app-of-apps Synced Healthy production-workloads Synced Healthy monitoring-stack Synced Healthy selfHeal: on prune: on
Full GitOps setup using App of Apps. AppProject with source/destination restrictions, per-application selfHeal policies, ArgoCD RBAC mapped to identity provider groups.
GitHub$ ./etcd/backup.sh Snapshot saved: etcd-20260519.db Verified: hash OK size: 4.2MB $ ./rbac-audit/find-clusteradmin.sh 2 bindings found — review needed
Operational runbooks for a live cluster — etcd backup/restore, certificate expiry checks, node isolation scripts, Falco alert triage playbook, and RBAC overprivilege auditing.
GitHub$ ip link show vxlan.calico vxlan.calico: mtu 1450 UP # Cilium tested — OOM killed host # reverted to Calico VXLAN podCIDR: 10.244.0.0/16
Calico VXLAN config, multi-tier NetworkPolicies, CoreDNS customization, ingress-nginx on bare metal. Honest post-mortem on a Cilium migration attempt that OOMed the control plane.
GitHubFrontend & Apps
Real-time cryptocurrency trading interface. Multi-order management, live portfolio tracking, authenticated session handling, and order book visualization via Bybit REST API.
GitHub
Full e-commerce storefront in React 18 and TypeScript. Full-viewport hero carousel, product grid, cart management, dark/light theme — 30+ components, zero UI library dependencies.
GitHub
Per-lesson gap analysis for Adrian Cantrill's AWS Pro course. Flags post-2021 service changes with OUTDATED / GAP / EXTRA tags. Supplements sorted inline so you study the delta at exactly the right time.
GitHubVerified Credentials
Get In Touch