tcplife [-h] [-T] [-t] [-w] [-s] [-p PID] [-D PORTS] [-L PORTS] [-4 | -6]
This tool traces TCP sessions that open and close while tracing, and prints a line of output to summarize each one. This includes the IP addresses, ports, duration, and throughput for the session. This is useful for workload characterisation and flow accounting: identifying what connections are happening, with the bytes transferred.
This tool works using the sock:inet_sock_set_state tracepoint if it exists, added to Linux 4.16, and switches to using kernel dynamic tracing for older kernels. Only TCP state changes are traced, so it is expected that the overhead of this tool is much lower than typical send/receive tracing.
Since this uses BPF, only the root user can use this tool.
CONFIG_BPF and bcc.
Print usage message.
Comma separated values output (parseable).
Include a timestamp column (seconds).
Include a time column (HH:MM:SS).
Wide column output (fits IPv6 addresses).
- -p PID
Trace this process ID only (filtered in-kernel).
- -L PORTS
Comma-separated list of local ports to trace (filtered in-kernel).
- -D PORTS
Comma-separated list of destination ports to trace (filtered in-kernel).
Trace IPv4 family only.
Trace IPv6 family only.
- Trace all TCP sessions, and summarize lifespan and throughput:
- Include a timestamp column, and wide column output:
# tcplife -tw
- Trace PID 181 only:
# tcplife -p 181
- Trace connections to local ports 80 and 81 only:
# tcplife -L 80,81
- Trace connections to remote port 80 only:
# tcplife -D 80
- Trace IPv4 family only:
# tcplife -4
- Trace IPv6 family only:
# tcplife -6
Time of the call, in HH:MM:SS format.
Time of the call, in seconds.
IP address family (4 or 6)
Local IP address.
Remote IP address.
Total transmitted Kbytes.
Total received Kbytes.
Lifespan of the session, in milliseconds.
This traces the kernel TCP set state function, which should be called much less often than send/receive tracing, and therefore have lower overhead. The overhead of the tool is relative to the rate of new TCP sessions: if this is high, over 10,000 per second, then there may be noticeable overhead just to print out 10k lines of formatted output per second.
You can find out the rate of new TCP sessions using "sar -n TCP 1", and adding the active/s and passive/s columns.
As always, test and understand this tools overhead for your types of workloads before production use.
This is from bcc.
Also look in the bcc distribution for a companion _examples.txt file containing example usage, output, and commentary for this tool.
Unstable - in development.
tcpaccept(8), tcpconnect(8), tcptop(8)