tcpretrans - Trace or count TCP retransmits and TLPs. Uses Linux eBPF/bcc. tcpretrans [-h] [-l] This traces TCP retransmits, showing address, port, and TCP state information, and sometimes the PID (although usually not, since retransmits are usually sent by the kernel on timeouts). To keep overhead very low, only the TCP retransmit functions are traced. This does not trace every packet (like tcpdump(8) or a packet sniffer). Optionally, it can count retransmits over a user signalled interval to spot potentially dropping network paths the flows are traversing. This uses dynamic tracing of the kernel tcp_retransmit_skb() and tcp_send_loss_probe() functions, and will need to be updated to match kernel changes to these functions. Since this uses BPF, only the root user can use this tool. CONFIG_BPF and bcc.

-h Print usage message.

-l Include tail loss probe attempts (in some cases the kernel may not complete the TLP send).

-c Count occurring retransmits per flow.

Trace TCP retransmits: # tcpretrans

Trace TCP retransmits and TLP attempts: # tcpretrans -l

TIME Time of the retransmit.

PID Process ID that was on-CPU. This is less useful than it might sound, as it may usually be 0, for the kernel, for timer-based retransmits.

IP IP address family (4 or 6).

LADDR Local IP address.

LPORT Local port.

T> Type of event: R> == retransmit, L> == tail loss probe.

RADDR Remote IP address.

RPORT Remote port.

STATE TCP session state.

RETRANSMITS Accumulated occurred retransmits since start.

Should be negligible: TCP retransmit events should be low (<1000/s), and the low overhead this tool adds to each event should make the cost negligible. This is from bcc. https://github.com/iovisor/bcc

Also look in the bcc distribution for a companion _examples.txt file containing example usage, output, and commentary for this tool.

Linux Unstable - in development. Brendan Gregg tcpconnect(8), tcpaccept(8)