mirror of
https://github.com/tailscale/tailscale.git
synced 2026-11-20 08:41:19 -08:00
bench.sh runs tailscale.com's TestTSWGBenchmark, which compares tailscaled in kernel mode against tailscaled in TUN mode (wireguard-go) on the same pair of natlab gokrazy VMs, and saves the results as JSON. Record the first results: bulk TCP throughput and CPU per byte are about the same (natlab caps throughput, and the kernel lacks outer UDP GSO/GRO), while the kernel receives about 1.7x the small packets per second at about 30% less CPU per packet and roughly halves latency under load. Also record the plan for a lean benchmark harness without natlab, whose userspace switch is expensive, and a flake seen when both peers restart at once. Updates #426 Signed-off-by: Brad Fitzpatrick <[email protected]> Change-Id: I69e696817f98e685576a30d766c42740d4d04810
23 lines
767 B
Bash
Executable File
23 lines
767 B
Bash
Executable File
#!/bin/sh
|
|
# SPDX-License-Identifier: GPL-2.0
|
|
#
|
|
# bench.sh runs TestTSWGBenchmark, which compares the tswg kernel data
|
|
# plane against wireguard-go on a pair of natlab gokrazy VMs, and writes
|
|
# the results as JSON to out/bench-<time>.json (or $TSWG_BENCH_OUT).
|
|
#
|
|
# Environment:
|
|
# TSWG_BENCH_CPUS virtual CPUs per VM (default 2)
|
|
# TSWG_BENCH_DUR length of each throughput run (default 10s)
|
|
# TSWG_BENCH_OUT where to write the JSON results
|
|
#
|
|
# usage: bench.sh [go test flags...]
|
|
|
|
set -eu
|
|
|
|
top=$(cd "$(dirname "$0")/.." && pwd)
|
|
mkdir -p "$top/out"
|
|
export TSWG_BENCH=1
|
|
export TSWG_BENCH_OUT="${TSWG_BENCH_OUT:-$top/out/bench-$(date -u +%Y%m%dT%H%M%SZ).json}"
|
|
echo "writing results to $TSWG_BENCH_OUT" >&2
|
|
exec "$top/scripts/vmtest.sh" 'TestTSWGBenchmark$' "$@"
|