net

Netto

Netto is an eBPF-based network monitoring tool for Linux that measures the CPU cost of the Linux network stack.

Features

• Measurement of the on-CPU time of the main networking entry points in kernel
• Breakdown of the NET_RX_SOFTIRQ entry point into the basic network functions by stack trace-based profiling of the kernel
• Low overhead
• Real time operation
• Minimum supported Linux version: 5.11 (BPF_MAP_TYPE_TASK_STORAGE)

Compilation

Building Netto as a standalone application is generally discouraged; users should instead use the provided Dockerfile to build a container image that will simplify deployment and management of the tool.

Build Netto into a container image (here tagged "netto") with:

docker build -t netto https://github.com/miolad/netto.git#perf-event-mmapable-array

Please note that the resulting image will need the CAP_BPF CAP_SYS_ADMIN capability, which under most setups means that the container should be run in --privileged mode.

If you'd still prefer to build the tool standalone you'll need:

• Latest Rust toolchain (stable or nightly)
• wasm_pack (for the web frontend)

Compile and run with:

cargo xtask run [--release]

Usage

$ ./netto -h
eBPF-based network diagnosis tool for Linux

Usage: netto [OPTIONS]

Options:
  -f, --frequency       Perf-event's sampling frequency in Hz for the NET_RX_SOFTIRQ cost breakdown     [default: 1000]
  -a, --address 
          Bind address for the web frontend [default: 0.0.0.0]
  -p, --port                 Bind port for the web frontend [default: 8080]
      --user-period   User-space controller update period in ms [default: 500]
  -l, --log-file         Path to a log file to which measurements are to be saved. If logging is enabled by providing this argument, any other form of web interface will be disabled
  -P, --prometheus                 Enable Prometheus logging in place of the web interface. The Prometheus-compatible endpoint will be available at `http://address:port`
  -h, --help                       Print help
  -V, --version                    Print version

Deployment

By default, Netto exposes the real time results as a Wasm-powered web page accessible on http://address:port. This interface is ideal for consuming the content as it is produced, but does not provide out-of-the-box support for storing the extracted metrics for delayed access. When real time monitoring of the host is not the primary concern, and to better integrate with existing metrics collection stacks (such as in data center environments), it is recommended to instead replace the built-in custom web frontend with a Prometheus-based exporter with the -P (or --prometheus) command line option.

In most cases the recommended way to run Netto involves a command similar to:

docker run --name netto --privileged --restart unless-stopped -d -p 8080:8080 netto -P

Or with the equivalent docker compose configuration entry:

services:
    netto:
        build: https://github.com/miolad/netto.git#perf-event-mmapable-array
        image: netto
        container_name: netto
        privileged: true
        command: -P
        restart: unless-stopped
    ...

Integration with Grafana

With the -P option selected, Netto can be targeted by the Prometheus scraper and referenced in Grafana directly. An example dashboard that showcases the major metrics collected by Netto is available as grafana_dashboard.json (which assumes that the variable "host" is used to differentiate between different monitored hosts through the Prometheus "job" label, to be configured in Prometheus' settings).

Repository structure

Folder	Contents
images	Images used on this README
netto	Main Rust binary crate
netto/src/bpf	BPF-C programs
metrics-common	Bridge Rust library crate for main and web-frontend
web-frontend	Rust WebAssembly frontend
www	Static files for the web frontend
xtask	Utility binary crate for managing coordination between the other crates