TCP port and network service monitoring
Monitor whether a database, gateway or custom network service is reachable on the expected port. RMON checks connectivity from selected locations and alerts when the service stops responding.
Check the service endpoint your applications depend on
A host can answer Ping while an important application port is unavailable. RMON resolves the service name, attempts the configured TCP connection and records whether a usable endpoint responded.
For services that expose a banner or accept a small request, the check can compare the returned text with the expected value. Multi-location checks show whether reachability differs by network path.
Monitor the right port
Check the actual service endpoint instead of relying only on host availability.
Validate the response
Send supported request data and compare a returned banner with the expected text.
Locate connectivity failures
Compare results and connection timing from different monitoring locations.
What this solution can cover
- DNS-name or IP-based TCP checks
- IPv4 and IPv6 connectivity
- Sequential attempts across resolved addresses
- Name-lookup and connection timing
- Optional request data
- Expected banner matching
- Retries and response-time thresholds
- Multi-location alerts and history
Create a TCP service check
- 01
Enter the endpoint
Choose the service DNS name or address and the TCP port it should accept.
- 02
Describe the response
Add request data and an expected banner when the protocol supports it.
- 03
Choose locations
Assign agents or regions that use the relevant network paths.
- 04
Set thresholds and alerts
Configure retries, timing limits, notification channels and a runbook.
Common questions about tcp port monitoring
Can RMON monitor any TCP port?
RMON can attempt a connection to the configured TCP endpoint and port, subject to normal network access.
Can a TCP check validate a service banner?
Yes. A check can send supported request data and compare the returned banner with expected text.
Does RMON support IPv6 TCP checks?
Yes. TCP checks can use IPv4 and IPv6 addresses returned for the service name.