TCP Ping
Measure connection latency at the TCP layer over the cellular connection.
Measure latency by opening TCP connections to a host instead of sending ICMP packets. Use it when firewalls filter ICMP, or to measure the latency your applications actually experience at the transport layer.
Parameters
| Parameter | Required | Description |
|---|---|---|
| Url | Yes | Host (and port) to reach. |
| Ping count | No | Number of connection attempts. Default: 10. |
| TCP Ping should fail | No | Invert the verdict. |
Plus the shared connection settings and DNS override (command.override_dns, command.dns).
Results and metrics
In the action report (result):
| Field | Description |
|---|---|
rtt_mean | Mean time to establish the TCP connection. |
packet_loss_pct | Percentage of failed attempts. |
ping_count | Number of attempts. |
dns_server | DNS servers used for resolution. |
duration | Total execution time in seconds. |
In Analytics, each run feeds the TCP ping measurement with rtt_seconds, packet_loss_percent, and duration_seconds, tagged by status, APN, DNS server, antenna, and access technology.
What's next?
Last updated on