System Architecture

Engineered for Massive Scale &
Zero-Trust Telemetry

A decoupled, highly reliable architecture designed to handle tens of thousands of simultaneous agent telemetry streams without lagging or dropping events.

01

Outbound-Only Go Agent (ops-agent)

mTLS / Port 443

Installed on edge Linux nodes via a single curl command. Operates strictly outbound, establishing mutual TLS with agent.opsvera.dev. Periodically streams encrypted heartbeats, CPU/RAM/Disk metrics, Docker status, and CrowdSec security events.

02

High-Throughput Ingestion & Queues

Laravel 12 + Horizon

Incoming telemetry packets are authenticated in microseconds and pushed to Redis queues. Horizon workers ingest metrics asynchronously, ensuring sub-second response times and zero bottlenecks during fleet-wide traffic spikes.

03

TimescaleDB Hypertable & Multi-Tenant RLS

PostgreSQL RLS

Time-series metrics are partitioned automatically into hypertable chunks for lightning-fast aggregation queries. Row-Level Security (RLS) is strictly enforced on all tables, providing mathematical isolation between tenants.

04

Vue 3 SPA & Real-Time Reverb WebSockets

Vue 3 + Reverb

The frontend control panel runs as a fast single-page app (SPA) at app.opsvera.dev. Live metric gauges, topology graph nodes, and alert toasts update seamlessly in real time without refreshing.

Service Domain Topology

opsvera.dev
Public Website & CMS
Public Web
app.opsvera.dev
Vue 3 SaaS Control Panel
Tenant SPA
api.opsvera.dev
Laravel 12 REST API
REST + Sanctum
agent.opsvera.dev
Go Agent Ingestion & mTLS
mTLS Ingress
docs.opsvera.dev
Developer Documentation & Guides
Technical Docs
Automated Inventory & Dependency Mapping

Visual Infrastructure Topology &
Instant Blast Radius Analysis

Connect assets and dependencies visually in the /topology/relations panel. When an upstream server, database, or proxy degrades, OpsVera pinpoints exactly which services are impacted.

Interactive Topology Canvas
source → target mapping
Cloudflare Edge & WAF
DNS • SSL • DDoS Protection
0ms Lag
Nginx Reverse Proxy & Load Balancer
CrowdSec Bouncer • Port 443
Healthy
Docker App Cluster
8 Containers • 0 Restart Loops
PostgreSQL Primary
TimescaleDB • Borg Backed

Precision Dependency & Relationship Engine

Easily configure which servers, databases, and microservices depend on one another directly from the UI. The API automatically rejects self-references and prevents invalid cross-workspace linkages.

Agent-to-Server Association
Seamlessly link ops-agent instances to logical server assets in one click.
Metadata-Driven Graphing
Dependencies are tracked via metadata.depends_on_asset_ids for instant topological querying.
Root-Cause Pinpointing
When an alert fires, immediately see if the fault originates in upstream DNS, network routing, or database lock contention.
Proactive Incident Guidance

Actionable Alarms &
Instant Remediation Runbooks

Never stare at cryptic monitoring alerts again. OpsVera pairs every infrastructure anomaly with root-cause context, severity classification, and step-by-step resolution advice.

CRITICAL • MEMORY 2m ago

PostgreSQL Buffer Saturation & Swap Spikes

Detected sudden buffer cache starvation and heavy swap I/O on primary DB node #03.

OpsVera Recommended Action:

Check for unindexed long-running analytical queries. Execute runbook: pg_terminate_idle_backends() and review shared_buffers allocation.

Slack & Webhook Dispatched View Runbook →
WARNING • SECURITY 14m ago

Coordinated SSH / WordPress Brute-Force Surge

Over 2,400 unauthorized login requests blocked across 12 client nodes in the last hour.

CrowdSec Automated Mitigation:

Malicious IP subnets have been propagated to all edge bouncers. Recommend enforcing MFA on client administrator accounts.

Fleet Bouncers Synchronized Inspect IPs →
INFO • BACKUP WATCHDOG 1h ago

BorgBackup Snapshot Size Delta Divergence

Daily snapshot on client workspace #acme-store is 45% smaller than baseline.

Recovery Runbook Trigger:

Verify excluded paths in backup profile. Automated test-restore script queued on isolated staging container.

Test Restore In Progress Check Logs →
Technical Documentation

OpsVera Agent Installation & Quickstart Guide

Deploy the ultra-lightweight ops-agent binary to any Linux server in under 60 seconds.

1

One-Liner Automated Installation

Run the following command on your target Linux machine as root or with sudo. The installer will automatically detect your OS distribution, architecture (x86_64 / aarch64), download the signed binary, configure systemd, and establish an outbound mTLS handshake.

curl -fsSL https://agent.opsvera.dev/install.sh | sudo bash -s -- --token=YOUR_WORKSPACE_TOKEN
2

Verify Agent Health & mTLS Handshake

Once installed, verify the service status with systemctl and run the diagnostic check:

// Check systemd daemon status
systemctl status opsvera-agent
// Run self-test and mTLS diagnostics
opsvera-agent --test-connection
3

System Requirements & Firewall Policy

Supported OS
  • Ubuntu 20.04, 22.04, 24.04 LTS
  • Debian 11 (Bullseye), 12 (Bookworm)
  • RHEL / AlmaLinux / Rocky Linux 8 & 9
  • Alpine Linux 3.16+
Minimal Footprint
  • RAM: < 15 MB resident memory
  • CPU: < 0.1% baseline overhead
  • Binary size: ~8.5 MB static Go ELF
  • Zero external shared libraries
Firewall & Ports
  • Inbound: 0 ports needed (None)
  • Outbound: TCP Port 443 (HTTPS)
  • Encrypted with TLS 1.3 + Client Cert
  • Safe through restrictive NAT/VPC
4

Backup Watchdog & Custom Health Hooks

OpsVera automatically detects active Borg and Restic repositories. To integrate custom cron backup scripts or trigger instant health pulses:

# Example: Sending a backup completion heartbeat from your cron script
/usr/local/bin/opsvera-agent backup-notify \
  --repo="borg://backup.corp.internal/var/backups" \
  --status=success \
  --bytes=42949672960 \
  --duration=312s
                    
Ready for Modern Operations?

Supercharge Your Infrastructure Operations Today

Join forward-thinking MSPs, cloud agencies, and DevOps teams who manage thousands of mission-critical Linux nodes effortlessly with OpsVera.