Octagotech's Server Infrastructure — VPS to Bare Metal, Fully Managed

The Brief: Managed, Not Just Hosted

Octagotech's business model requires reliable, performant hosting infrastructure for their clients. They had existing servers but no systematic management process. Backups were inconsistent.

Network configuration had grown organically without a security review. When attacks occurred, there was no structured response — incidents were handled reactively, often after impact had already occurred. They needed a managed infrastructure partner: not a hosting company that rents them servers and disappears, but a team that treats their infrastructure as a managed service.

That distinction — managed versus hosted — is the one that mattered most.

VPS and Bare Metal: A Layered Architecture

Octagotech's workloads required both VPS instances for flexible, rapidly-provisioned capacity and dedicated bare metal servers for workloads requiring predictable performance without hypervisor overhead. We standardised the configuration across both tiers: consistent OS hardening baselines, unified monitoring agents, standardised network policy rules, and a single management interface for both environments. cPanel and WHM were deployed on the appropriate server tiers to give Octagotech's clients self-service access to their hosting accounts, while all underlying server management remained under our control. This separation — client-accessible control panels sitting above infrastructure that clients cannot touch — is the right architecture for a managed hosting business.

It eliminates the risk of a client misconfiguration taking down the underlying server.

Multi-Layer Backup Architecture

We designed and implemented a backup system with three independent recovery layers. The first layer is automated daily incremental backups to a separate local storage volume on each server — fast to restore, low cost, useful for the most common scenario: accidental deletion or corruption. The second layer is nightly full backups to an off-site storage location with a different geographic footprint — protection against physical server failure or datacenter-level event.

The third layer is weekly snapshots retained for thirty days — providing a longer recovery window for scenarios where damage or compromise was not immediately detected. Each layer has a documented restore procedure that has been tested under realistic conditions. A backup system that has never been tested is not a backup system.

Routing, Network Security, and Attack Response

Network configuration was one of the first things we reviewed and rebuilt. Firewall rules had accumulated over time without a consistent policy. We audited every inbound and outbound rule, removed unnecessary open ports, and implemented a deny-by-default policy with explicit allowlists for required services.

Rate limiting was applied at the network layer to mitigate brute force and scanning attempts. DDoS mitigation rules were configured to detect and automatically null-route attack traffic before it reached application layer. When attacks occur — and they do, regularly, against any internet-facing infrastructure — our team receives an alert, reviews the attack signature, and implements the most effective mitigation within the client's budget and infrastructure constraints.

We have responded to volumetric attacks, application-layer attacks targeting cPanel, and SSH brute force campaigns. In each case, the response time from alert to mitigation was under fifteen minutes.

Ongoing Management and What It Means in Practice

Managed infrastructure means we are accountable for uptime, not just availability of the physical server. We monitor every service, every backup job, and every network anomaly. We apply security patches within forty-eight hours of a critical CVE publication.

We review server health weekly and flag anything that warrants attention before it becomes a problem. Octagotech's clients experience this as exceptional reliability — they do not see the infrastructure work that makes it possible. That invisibility is exactly what good managed infrastructure should feel like.