tcptracer-bpfcc8
Trace tcp established connections. uses linux ebpf/bcc.
Usage
tcptracer [-h]
[-v]
[-t]
[-p
PID]
[-N
NETNS]
[--cgroupmap
MAPPATH]
[--mntnsmap
MAPPATH]
[{-4 | -6}]
Overview
This tool traces established TCP connections that open and close while tracing, and prints a line of output per connect, accept and close events. This includes the type of event, PID, IP addresses and ports.
This tool works by using kernel dynamic tracing, and will need to be updated if the kernel implementation changes. Only established TCP connections are traced, so it is expected that the overhead of this tool is rather low.
Since this uses BPF, only the root user can use this tool.
Requirements
CONFIG_BPF and bcc.
Options
-
-h -
Print usage message.
-
-v -
Print full lines, with long event type names and network namespace numbers.
-
-t -
Include timestamp on output
-
-p PID -
Trace this process ID only (filtered in-kernel).
-
-N NETNS -
Trace this network namespace only (filtered in-kernel).
-
--cgroupmap MAPPATH -
Trace cgroups in this BPF map only (filtered in-kernel).
-
--mntnsmap MAPPATH -
Trace mount namespaces in the map (filtered in-kernel).
-
-4 -
Trace IPv4 family only.
-
-6 -
Trace IPv6 family only.
Examples
Trace all TCP established connections:
|
# |
|---|---|
Trace all TCP established connections with verbose lines:
|
# tcptracer -v |
Trace PID 181 only:
|
# tcptracer -p 181 |
Trace connections in network namespace 4026531969 only:
|
# tcptracer -N 4026531969 |
Trace a set of cgroups only (see special_filtering.md from bcc sources for more details):
|
# tcptracer --cgroupmap /sys/fs/bpf/test01 |
Trace IPv4 family only:
|
# tcptracer -4 |
Trace IPv6 family only:
|
# tcptracer -6 |
Fields
TYPE
|
Type of event. In non-verbose mode: C for connect, A for accept, X for close. |
|---|---|
PID
|
Process ID |
COMM
|
Process name |
IP
|
IP address family (4 or 6) |
SADDR
|
Source IP address. |
DADDR
|
Destination IP address. |
SPORT
|
Source port. |
DPORT
|
Destination port. |
NETNS
|
Network namespace where the event originated. |
Overhead
This traces the kernel inet accept function, and the TCP connect, close, and set state functions. However, it only prints information for connections that are established, so it shouldn't have a huge overhead.
As always, test and understand this tools overhead for your types of workloads before production use.
Source
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.
OS
Linux
Stability
Unstable - in development.
Releases
4 releases since Debian 10 (Buster); the text changed 2 times.
--cgroupmap since Debian 11 (Bullseye);
--mntnsmap since Debian 11 (Bullseye);
-4 since Debian 12 (Bookworm);
-6 since Debian 12 (Bookworm);
-t since Debian 12 (Bookworm).