Like many people in IT, I’ve been running a home lab for several years. My home lab has become progressively more complicated over the years as I’ve layered in new technologies that I want to explore and added new services to my home network.

Hardware#

My home lab runs mainly on low-power Dell OptiPlex ultra-small-form-factor (USFF) hardware. I wanted this lab to be always-on - so small, quiet, and energy-efficient hardware is a must-have. The notable exception is my NAS, a 2U Dell PowerEdge rackmount server with an array of 3.5" SAS spinning disks striped in a large ZFS RAID-Z2 pool.

Hypervisors#

For several years, I ran VMware ESXi mostly as a learning opportunity. Over time I found that the overhead of running ESXi on low-spec commodity hardware was too great, and decided to shift my workloads over to Proxmox VE.

Services#

Each physical system in my homelab runs the Debian-based Proxmox VE hypervisor. The hypervisors host a mix of LXC-based containers and QEMU virtual machines. From top to bottom:

  • HAProxy
    • Two very lightweight LXD containers serving as a virtual network load balancer
    • HAProxy acts as a TCP and HTTP load balancer
    • Keepalived manages virtual IPs that failover between the two containers
  • MariaDB Galera Cluster
    • Three Debian LXD containers running MariaDB with Galera multi-master replication
    • Clients access the MariaDB cluster using a VIP and server pool managed by the HAProxy cluster
    • I wrote a monitoring daemon in Go that provides HTTP healthcheck capabilities for HAProxy to determine the health of each MariaDB instance
  • Ceph
    • Each Proxmox host runs a Ceph monitor
    • Each host also hosts a Ceph OSD on an internal 2.5" SATA disk
  • Active Directory
    • Three MSDN Server 2019 Core VMs across each host provide Active Directory services
  • Kubernetes
    • 3 control plane nodes with taints
    • 1 worker node
    • Workloads on Kubernetes are able to leverage the external MariaDB and Ceph clusters for persistence

In addition to these clustered services, I also deployed a few standalone services for better management and visibility across my environment:

  • Chef Infra Server
    • Manages automation for VMs, LXD containers, and physical hosts
    • I’ve also open-sourced a few of the cookbooks I created to manage my lab systems
  • Elastic ELK Stack
    • Logstash takes syslog data and Kubernetes logs from Filebeat and indexes them into Elasticsearch
    • A Kibana instance running on Kubernetes provides search and visualization capabilities
  • InfluxDB
    • Aggregates time-series monitoring data from telegraf clients on Linux
    • Provides data to a Grafana instance running on Kubernetes