CASE STUDY · DEVOPS / INFRASTRUCTURE5 min read

How we built a server management dashboard that brought VM provisioning from hours to minutes

A cloud infrastructure team was provisioning and managing virtual machines manually via CLI and vendor consoles. We replaced it with a unified management platform integrating Proxmox and KT Cloud.

Industry

Cloud Infrastructure

Timeline

3 Months

My Role

Lead Developer & Architect

Integrations

Proxmox + KT Cloud

THE PROBLEM

Managing 50+ VMs across two platforms entirely through CLI and vendor dashboards

The infrastructure team was managing a growing fleet of virtual machines across two separate environments — an on-premise Proxmox cluster and KT Cloud. Every task required logging into a different console, running CLI commands, or using two completely different vendor interfaces.

Provisioning a new VM took 20–40 minutes per server. There was no unified view of resource usage across both environments. Non-technical team members could not manage anything without developer assistance. Billing data from both providers had to be reconciled manually.

→No unified view of VMs across Proxmox and KT Cloud

→VM provisioning was slow, manual, and error-prone

→Monitoring scattered across multiple tools

→Non-technical staff completely locked out

→No audit log of who changed what and when

→Resource usage had no alerting — issues caught too late

THE SOLUTION

A single dashboard for your entire infrastructure — on-premise and cloud

All virtual machines from both Proxmox and KT Cloud appear in one dashboard — status, CPU, RAM, disk usage, uptime, and IP all visible at a glance. VM lifecycle management (start, stop, restart, suspend, delete) is available directly from the dashboard with full audit logging on every action.

Unified VM dashboard — all environments in one view

One-click provisioning — 30 mins reduced to under 3 minutes

Real-time monitoring — CPU, RAM, disk polled every 60 seconds

Configurable alerts — email + in-app when thresholds crossed

Full audit log — every action permanently recorded

Role-based access — engineers, ops staff, and management

Provider abstraction — new cloud providers addable in days

TECHNICAL HIGHLIGHTS

Why we built it this way

API abstraction layer across two providers

Proxmox and KT Cloud have completely different APIs. We built an internal provider interface that standardises all VM operations — create, list, start, stop, delete — behind a single internal API. Adding a third cloud provider in future requires only implementing this interface, not rewriting the entire dashboard.

Background polling with smart caching

Querying 50+ VMs in real-time on every page load would have been too slow and hammered the provider APIs. We implemented a background job that polls all VMs every 60 seconds and caches results in Redis. The dashboard reads from cache — instant load — and a manual refresh button triggers an immediate poll.

Audit log for every infrastructure action

Every VM action — start, stop, provision, delete — is recorded with the operator's identity, timestamp, parameters, and result. This audit log is immutable (append-only) and accessible to admins. In a regulated infrastructure environment, this is not optional — it's essential.

Alert threshold system without a monitoring platform

Rather than integrating a heavy external monitoring platform (Datadog, New Relic — expensive for this scale), we built a lightweight threshold engine in-app. Each VM has configurable alert rules. The background polling job evaluates thresholds on every cycle and fires alerts when crossed.

THE RESULTS

Live, growing, and trusted

VM provisioning time reduced from 30 minutes to under 3 minutes

Unified dashboard — all VMs from Proxmox and KT Cloud in one view

Real-time monitoring with configurable alerts across all servers

Full audit log — every infrastructure action permanently recorded

Non-technical staff can monitor infrastructure without developer help

Role-based access — right visibility for every team member

Provider abstraction layer — new cloud providers addable in days

TECH STACK

Node.jsExpressVue.jsMySQLRedisDockerProxmox APIKT Cloud APIBackground JobsREST APIJWT AuthRole-based Access ControlNginxUbuntu Server

Managing infrastructure across multiple environments manually?

We build internal tools and dashboards that give your team full control without the CLI. Let's talk.

Discuss your project

BACK TO PORTFOLIO

View All Projects