Why Open Source Services Support Plans Case Studies Support Details Book Consultation
Case Study · Professional Services

From fragmented SaaS
to a sovereign,
self-hosted office.

A 45-person consulting firm across three offices replaced Microsoft 365, a shared NAS, and scattered cloud tools with a fully self-hosted, centrally managed open-source stack — delivered in eight weeks.

Industry Professional Services
Team Size 45 employees · 3 offices
Delivery Time 8 weeks
Stack 100% open source
The Challenge

A growing firm outgrowing its tools — and its budget.

The client was spending heavily on SaaS subscriptions while struggling with fragmented data, no central authentication, and zero visibility into who was accessing what. Each office operated independently with no shared infrastructure.

No centralised identity

Each user had separate credentials for file storage, email, and internal tools. Offboarding an employee took hours and often left orphaned accounts.

Data jurisdiction risk

Files in OneDrive, email on Exchange Online, and CRM in a US-hosted SaaS — none of it under the firm's legal control or their country's jurisdiction.

Flat network with no segmentation

All devices — workstations, servers, printers, and guest Wi-Fi — shared the same subnet with no firewall between them.

No backup or disaster recovery

Data existed in a single location. One hardware failure or ransomware event would have been catastrophic with no tested recovery path.

Before — Existing Stack

Microsoft 365 Business
Email + OneDrive + Teams
$2,700 / mo
Dropbox Business
File sharing across offices
$560 / mo
Separate AD per office
No sync, manual GPO
Fragmented
Unmanaged switches
Single flat subnet
No segmentation
No monitoring
Failures discovered by users
Blind
No backup strategy
Single copy, no offsite
No DR
Total monthly SaaS spend $3,260+
Discovery & Assessment

Understand everything
before touching anything.

We began with a structured two-week discovery process — no assumptions, no templates. Every finding shaped the final architecture.

01

Network Audit

Topology mapping, device inventory, traffic analysis, and firewall rule review across all three sites.

02

Workload Analysis

Identifying which services were business-critical, storage growth rate, email volume, and VPN usage patterns.

03

Compliance Review

Data residency requirements, access control obligations, and audit trail needs for the firm's regulatory context.

04

Proposal & Sign-off

Architecture document, bill of materials, phased timeline, and cost model presented to the client for approval.

Storage in use
2.4 TB

Spread across OneDrive, local NAS, and personal laptops with no deduplication or lifecycle policy.

Unmanaged devices
38

Laptops, workstations, and printers with no MDM, no patch policy, and mixed OS versions.

Open firewall rules
14

Overly permissive inbound and inter-VLAN rules that granted unnecessary lateral movement paths.

Identity systems
3

Three separate directory systems — one per office — with no federation or password sync.

Last backup verified
Never

Backups existed in theory. No restore test had ever been performed. Two drives were found to have failed silently.

Potential savings
~$38k

Annual SaaS subscription savings identified after mapping every paid tool to a self-hosted equivalent.

Also Assessed

Odoo ERP — Business Workflow Assessment

Beyond infrastructure, we conducted a structured review of how the firm runs its core business operations — including project management, invoicing, client tracking, and procurement. We mapped each manual process to an Odoo module, identified gaps, and configured the platform to match how the team actually works — rather than forcing the team to adapt to a generic ERP template. The result was a deployed, trained, and live Odoo instance within the same eight-week window.

Solution Architecture

A full-stack, sovereign office infrastructure.

Every layer designed for redundancy, security, and operational simplicity. Open source throughout.

Internet
pfSense Firewall
Dual WAN Failover  ·  Suricata IDS/IPS  ·  WireGuard VPN Gateway
WireGuard VPN
Remote Employees & Managers
WireGuard peers · encrypted tunnel
Managed Switch — 802.1Q Trunk
VLAN 10
Management
10.10.10.0/24
VLAN 20
Employees
10.10.20.0/24
VLAN 30
Servers
10.10.30.0/24
VLAN 40
IoT / Printers
10.10.40.0/24
Proxmox Node 1
Primary
Proxmox Node 2
Secondary
Proxmox Node 3
Quorum / DR
HA Cluster — Corosync · Ceph Storage · Live Migration
VMs & LXC Containers
OpenLDAP Nextcloud Collabora Online Stalwart Mail Odoo ERP Wazuh SIEM Prometheus Grafana BorgBackup WireGuard

Proxmox VE Cluster

3-node HA cluster running all workloads as VMs and LXC containers. Live migration and automatic failover with no single point of failure.

Proxmox VE 8HA ClusterCorosyncCeph Storage

pfSense Firewall

Perimeter firewall with dual WAN failover, strict inter-VLAN routing rules, Suricata IDS/IPS, and WireGuard site-to-site VPN for all three offices.

pfSense CESuricata IDSDual WANVLAN Routing

OpenLDAP — Central Identity

Single directory for all employees across all sites. One account, one password — works for email, Nextcloud, VPN, and Wi-Fi (802.1X). Offboarding is one command.

OpenLDAPLDAP TLS802.1X AuthGroup Policies

Nextcloud — File & Collaboration

Replaces OneDrive and Dropbox. File sync, document editing via Collabora Online, team calendars, contacts, and video calls — all authenticated via LDAP.

Nextcloud 29Collabora OnlineLDAP AuthS3 Backup

Stalwart — Self-hosted Email

Full mail server handling inbound and outbound email for all staff. LDAP-authenticated IMAP/SMTP, DKIM/DMARC/SPF configured, spam filtering enabled.

Stalwart MailDKIM / DMARCLDAP AuthAnti-spam

WireGuard VPN

Site-to-site tunnels connecting all three offices into one unified network. Remote employees connect via individual WireGuard peers — full access to internal resources.

WireGuardSite-to-siteSplit TunnelMFA via LDAP

Odoo ERP — Business Operations

We conducted a full workflow assessment of the client's business processes — accounting, HR, CRM, and procurement — then configured and deployed Odoo tailored to their exact operational model. Staff were trained before go-live.

Odoo 17Workflow AssessmentCRMAccountingLDAP Auth
Delivery Timeline

Eight weeks,
zero production downtime.

Migration was executed in parallel with the existing SaaS stack. Employees cut over service by service — no big-bang migration, no weekend outages.

Phase 01

Foundation & Network

Weeks 1 – 2
Proxmox 3-node cluster installed and validated with HA
Managed switches deployed, 802.1Q VLANs configured
pfSense deployed with dual WAN, VLAN routing, and firewall rules
WireGuard site-to-site tunnels established between all 3 offices
OpenLDAP provisioned with all 45 user accounts and OUs
DNS internal zones configured for all services
Phase 02

Core Services

Weeks 3 – 5
Nextcloud deployed with LDAP auth, 8 TB Ceph-backed storage
Collabora Online integrated for live document editing
2.4 TB migrated from OneDrive and local NAS with zero data loss
Stalwart mail server provisioned with SPF, DKIM, DMARC
Three years of email history migrated via IMAP sync
MX records cut over with parallel delivery during transition
Phase 03

Security & Observability

Weeks 6 – 8
Wazuh SIEM deployed, agents installed on all servers and endpoints
Prometheus + Grafana dashboards for infrastructure metrics
Uptime Kuma configured with PagerDuty-style SMS alerting
BorgBackup configured with encrypted daily snapshots
Offsite replication to encrypted object storage in local data centre
Full disaster recovery test executed — 4-hour RTO verified
Ongoing Managed Support

Delivered. Then watched,
maintained, and improved.

Post-deployment the client moved to Hanshala's Pro support plan. Their infrastructure is managed continuously — no internal IT hire needed.

24/7 Infrastructure Monitoring

Every service, VM, VLAN, and WAN link monitored continuously. Alerts fire before users notice a problem.

  • Uptime, latency, and disk usage tracking across all nodes
  • pfSense WAN failover tested monthly
  • Grafana dashboard shared with client management
  • Incident response within 1 hour for critical alerts

Security & Patch Management

Security patches applied on a tested schedule. Wazuh SIEM actively monitored for threats and anomalies.

  • OS and application updates on monthly maintenance windows
  • CVE tracking for all deployed software components
  • Wazuh alerts reviewed daily, escalated when needed
  • Quarterly penetration testing and firewall rule review

Backup & Disaster Recovery

Automated encrypted backups run daily. Restore tests run quarterly. Recovery time objective maintained at under 4 hours.

  • BorgBackup daily snapshots with 90-day retention
  • Offsite replication to geographically separate location
  • Quarterly restore drill with documented results
  • Ransomware-resilient — air-gapped backup generation

User & Identity Management

New hires provisioned same day. Leavers fully offboarded across every service in minutes — not hours.

  • LDAP account provisioning within 2 hours of request
  • Group-based access control for Nextcloud and email
  • Quarterly access review report delivered to management
  • Self-service password reset portal for employees
Outcomes

Measurable results,
six months later.

Numbers reported by the client six months post-deployment compared to the pre-migration baseline.

$38k
Annual SaaS savings
99.96%
Infrastructure uptime
<4h
Verified DR recovery time
0
Security incidents post-go-live

Full data sovereignty achieved

All business data — email, files, calendars — now resides on hardware the client owns, in their country, under their jurisdiction.

Single identity, every service

One LDAP account authenticates email, file storage, VPN, and Wi-Fi. New employees are productive within hours of joining.

Segmented, auditable network

VLANs isolate employee, server, and IoT traffic. Every inter-VLAN access is logged. The flat network that allowed unrestricted lateral movement is gone.

Backup actually works

Daily encrypted snapshots, tested quarterly. The first successful restore test — ever — was completed in week 8 of the project with a 3h 42m RTO.

No internal IT hire needed

The managed support contract covers everything the client would have hired a sysadmin for — at a fraction of the cost, with deeper expertise.

Vendor lock-in eliminated

Every component is open source and standards-based. The client can move, fork, or self-manage any service at any time — no contract holds them hostage.

Your Turn

Does this look like
your organisation?

Every engagement starts with a free assessment. We map what you have, identify the gaps, and propose what should be built — before you commit to anything.