Ticketing · Observability

Muchticket

Full visibility across a high-concurrency ticketing platform — before any user notices a problem.

muchticket.com
Client
Muchticket
Industry
Ticketing · High concurrency
Services
Observability · SRE · Alerting
Platforms
Prometheus · Grafana · BlackBox · Telegram
The challenge

Blind spots during the moments that matter most

Muchticket sells event tickets across Latin America and Europe, with virtual queues that must hold up when a hundred thousand fans hit "buy" at the same second. Their on-premise infrastructure had no centralized monitoring, no structured alerting, and zero visibility into server health until something broke.

The IT team found out about incidents through user complaints — not metrics. By then, tickets were already lost and the on-sale was already in chaos. They needed a system that detected problems before any end user noticed, and got the right people alerted in seconds, not minutes.

The solution

A centralized observability stack with instant alerting

  1. We deployed Prometheus as the central metrics engine, scraping every server, queue worker, and process on a fixed interval. This gives the team a continuous time-series record of the entire infrastructure. Grafana sits on top, translating raw metrics into real-time dashboards and historical trend views — a single pane of glass for the full on-premise environment.

  2. BlackBox Exporter probes every public HTTP endpoint from the outside, measuring uptime and response latency from the perspective that matters: the end user's. If a service degrades or goes down, BlackBox detects it in the next scrape cycle — before any ticket buyer notices.

  3. A Telegram Bot delivers structured alerts to the on-call channel in seconds. Each message includes the affected host, the metric name, the current value, and the threshold crossed — exactly the context needed to act immediately, without digging through dashboards while an on-sale burns.

Architecture
On-Premise Infrastructure
Web Server
Queue Worker
DB Server
App Server N
metrics scrape (pull)
Prometheus Time-series storage & alerting
Grafana Dashboards & trends
BlackBox Exporter HTTP endpoint probing
Telegram Bot On-call alerts in seconds
Prometheus pulls metrics from every node. BlackBox probes endpoints from the outside. Alerts fire before any end user notices degradation.
Tech stack

Tools we used

  • Prometheus Time-series metrics engine
  • Grafana Real-time dashboards & alerts
  • BlackBox Exporter External HTTP endpoint probing
  • Telegram Bot Instant on-call alerting
  • Node Exporter Server-level OS metrics
  • Alertmanager Alert routing & deduplication
Results

From blind to fully instrumented

100% server coverage
<30s alert delivery time
0 missed incidents
24/7 continuous monitoring
"Pluralty transformed how we operate. Before, we found out about incidents through user complaints. Now our system tells us before anyone notices. That shift alone changed everything for our on-sale events."
Luis María Springolo IT Coordinator / SRE, Muchticket
Flying blind on your infrastructure?

Let's fix that — starting with a free assessment

We'll map your monitoring blind spots and show you exactly what to instrument first.

Get your free assessment