Monitor any TCP service with reliable port checks
TCP Port Monitoring
Databases, caches, message queues, mail servers. If it listens on a port, upti.my can monitor it. Get alerted the moment a connection fails, before your application errors cascade.
Your services are more than HTTP endpoints
Modern infrastructure runs on dozens of TCP services. Your PostgreSQL database, Redis cache, RabbitMQ queue, and SMTP server all need monitoring. But most tools only check HTTP. When your database port stops accepting connections, you find out from failed requests. Not from your monitoring.
Database Monitoring
PostgreSQL, MySQL, MongoDB, Redis port checks.
Service Discovery
Verify services are listening before traffic arrives.
Fast Detection
30-second checks catch failures immediately.
Send/Expect Patterns
Validate protocol handshakes beyond just connection.
Common TCP monitoring gaps
Many teams skip TCP monitoring entirely, assuming their databases and services "just work." Here's what goes wrong:
- ✗Connection pool exhaustion where max connections are reached and new requests fail
- ✗Firewall rule changes where security updates silently block ports
- ✗Service crashes where the process dies but container stays "healthy"
- ✗Network partitions where the service runs but is unreachable from other zones
Simple, reliable TCP health checks
upti.my connects to your TCP ports every 30 seconds from multiple regions. We detect connection refusals, timeouts, and handshake failures. Get alerts instantly via Slack, Discord, or webhooks.
- ✓Monitor any port on any host
- ✓Optional send/expect patterns for protocol validation
- ✓Configurable connection timeouts
- ✓Multi-region checks to distinguish local vs global issues
Getting Started
Set up TCP monitoring in the dashboard:
- 1
Create a TCP Monitor
Select TCP as your monitor type. Enter the host and port you want to check.
- 2
Set Connection Timeout
Choose how long to wait for a connection. 5-10 seconds works for most services.
- 3
Add Send/Expect Pattern (Optional)
For protocol validation, send a string after connecting and verify the response. Example: send PING to Redis, expect PONG.
- 4
Select Check Regions
Pick multiple regions to distinguish between local and global connectivity issues.
Frequently Asked Questions
What TCP services can I monitor?
You can monitor any TCP-based service including databases (PostgreSQL, MySQL, MongoDB), message queues (Redis, RabbitMQ), mail servers (SMTP, IMAP), custom application ports, and any other service that accepts TCP connections.
How does TCP monitoring differ from HTTP monitoring?
TCP monitoring checks if a port accepts connections at the transport layer, while HTTP monitoring validates the full application response. Use TCP for services that don't speak HTTP, like databases, caches, or custom protocols.
Can I verify data after connecting?
Yes, upti.my supports TCP checks with optional send/expect patterns. You can send a string after connecting and verify the response matches expected data, useful for protocol handshakes.
What timeout should I set for TCP checks?
We recommend 5-10 seconds for most services. Local or well-provisioned services can use shorter timeouts (2-3s), while services with cold-start behavior may need longer (15-30s).
Related Topics
Ready to start monitoring?
Set up your first health check in under 2 minutes. No credit card required.