cmd/derpprobe,prober: add ability to perform continuous queuing delay measurements against DERP servers

This new type of probe sends DERP packets sized similarly to CallMeMaybe packets
at a rate of 10 packets per second. It records the round-trip times in a Prometheus
histogram. It also keeps track of how many packets are dropped. Packets that fail to
arrive within 5 seconds are considered dropped.

Updates tailscale/corp#24522

Signed-off-by: Percy Wegmann <percy@tailscale.com>
This commit is contained in:
Percy Wegmann
2024-12-16 23:05:46 -06:00
committed by Percy Wegmann
parent 6ae0287a57
commit 00a4504cf1
8 changed files with 429 additions and 55 deletions

View File

@@ -18,19 +18,21 @@ import (
)
var (
derpMapURL = flag.String("derp-map", "https://login.tailscale.com/derpmap/default", "URL to DERP map (https:// or file://) or 'local' to use the local tailscaled's DERP map")
versionFlag = flag.Bool("version", false, "print version and exit")
listen = flag.String("listen", ":8030", "HTTP listen address")
probeOnce = flag.Bool("once", false, "probe once and print results, then exit; ignores the listen flag")
spread = flag.Bool("spread", true, "whether to spread probing over time")
interval = flag.Duration("interval", 15*time.Second, "probe interval")
meshInterval = flag.Duration("mesh-interval", 15*time.Second, "mesh probe interval")
stunInterval = flag.Duration("stun-interval", 15*time.Second, "STUN probe interval")
tlsInterval = flag.Duration("tls-interval", 15*time.Second, "TLS probe interval")
bwInterval = flag.Duration("bw-interval", 0, "bandwidth probe interval (0 = no bandwidth probing)")
bwSize = flag.Int64("bw-probe-size-bytes", 1_000_000, "bandwidth probe size")
bwTUNIPv4Address = flag.String("bw-tun-ipv4-addr", "", "if specified, bandwidth probes will be performed over a TUN device at this address in order to exercise TCP-in-TCP in similar fashion to TCP over Tailscale via DERP. We will use a /30 subnet including this IP address.")
regionCode = flag.String("region-code", "", "probe only this region (e.g. 'lax'); if left blank, all regions will be probed")
derpMapURL = flag.String("derp-map", "https://login.tailscale.com/derpmap/default", "URL to DERP map (https:// or file://) or 'local' to use the local tailscaled's DERP map")
versionFlag = flag.Bool("version", false, "print version and exit")
listen = flag.String("listen", ":8030", "HTTP listen address")
probeOnce = flag.Bool("once", false, "probe once and print results, then exit; ignores the listen flag")
spread = flag.Bool("spread", true, "whether to spread probing over time")
interval = flag.Duration("interval", 15*time.Second, "probe interval")
meshInterval = flag.Duration("mesh-interval", 15*time.Second, "mesh probe interval")
stunInterval = flag.Duration("stun-interval", 15*time.Second, "STUN probe interval")
tlsInterval = flag.Duration("tls-interval", 15*time.Second, "TLS probe interval")
bwInterval = flag.Duration("bw-interval", 0, "bandwidth probe interval (0 = no bandwidth probing)")
bwSize = flag.Int64("bw-probe-size-bytes", 1_000_000, "bandwidth probe size")
bwTUNIPv4Address = flag.String("bw-tun-ipv4-addr", "", "if specified, bandwidth probes will be performed over a TUN device at this address in order to exercise TCP-in-TCP in similar fashion to TCP over Tailscale via DERP; we will use a /30 subnet including this IP address")
qdPacketsPerSecond = flag.Int("qd-packets-per-second", 0, "if greater than 0, queuing delay will be measured continuously using 260 byte packets (approximate size of a CallMeMaybe packet) sent at this rate per second")
qdPacketTimeout = flag.Duration("qd-packet-timeout", 5*time.Second, "queuing delay packets arriving after this period of time from being sent are treated like dropped packets and don't count toward queuing delay timings")
regionCode = flag.String("region-code", "", "probe only this region (e.g. 'lax'); if left blank, all regions will be probed")
)
func main() {
@@ -45,6 +47,7 @@ func main() {
prober.WithMeshProbing(*meshInterval),
prober.WithSTUNProbing(*stunInterval),
prober.WithTLSProbing(*tlsInterval),
prober.WithQueuingDelayProbing(*qdPacketsPerSecond, *qdPacketTimeout),
}
if *bwInterval > 0 {
opts = append(opts, prober.WithBandwidthProbing(*bwInterval, *bwSize, *bwTUNIPv4Address))
@@ -107,7 +110,7 @@ func getOverallStatus(p *prober.Prober) (o overallStatus) {
// Do not show probes that have not finished yet.
continue
}
if i.Result {
if i.Status == prober.ProbeStatusSucceeded {
o.addGoodf("%s: %s", p, i.Latency)
} else {
o.addBadf("%s: %s", p, i.Error)