netwatch ~ ~/cloud.tsx
main·v0.4.2
nw
netwatchcloud
// 01 · what is this
> linux fleet
monitoring,
without the bloat;
A 5 MB Rust agent and a hosted dashboard. Drop it on any Linux host with one curl command — host shows up in the dashboard within 15 seconds.
5MB
agent_size
<2min
ttf_host
0
config_files
15s
cadence
app.netwatch.cloud/fleetsample fleet
// 27-second walk-through · sample fleet · loops · no audio▶ playing
// 02 · the dashboard
Watch it work.
five hosts, one screen, real time.
Real web app, recorded against a representative fleet. Sample hosts, fixture metrics — the UI and interactions are exactly what ships.
app.netwatch.cloud/fleetlive demo
00:02fleet overview
every host on one screen, ranked by health
00:08top offenders
hosts above your thresholds, worst first
00:14live charts
rx/tx and cpu/load streaming at 15s cadence
00:19host detail
click any host for 72-hour replay
/* 01 */
no agents that need agents.
5 MB Rust binary. Zero runtime deps. Static linking. The agent is the agent.
/* 02 */
no YAML, no PromQL, no rules.
Built-in alerts cover the failures that actually happen. You configure thresholds, not a query language.
/* 03 */
no enterprise observability tax.
Free while we grow. Paid tiers may arrive later for higher-volume use; existing features stay free.
// 03 · what you get
fleet.dashboard
Every host, ranked by health.
CPU, memory, load, latency, packet loss — every host you run, sortable, filterable. Click into any host’s last 72 hours.
db-replica-01cpu 21%mem 64%ok
edge-router-02cpu 8%mem 41%ok
queue-worker-03cpu 74%mem 88%warn
web-prod-01cpu 56%mem 60%ok
web-prod-02cpu 0%mem 0%down
alerts.rules
Pinpoint, not page.
Slack + email when something actually breaks. Auto-resolve when it stops.
agent.binary
5 MB. Zero deps.
Static Rust binary. systemd service. No root. Self-updating.
probes.network
Find the broken hop.
Gateway latency, DNS resolution, packet loss — sampled every 30s.
history.metric
72-hour replay.
Interactive charts on every metric, every host. No retention surprises.
// 04 · where this fits
There are great tools at both extremes.
Heavyweight observability platforms. Bare-bones self-hosted pingers. NetWatch sits in between — host-level network monitoring that just works, with zero ops overhead.
field
NetWatch
Datadog
Uptime Kuma
PRTG
setup_time
2 min
30+ min
15+ min
1+ hr
agent_footprint
5 MB
~800 MB
external
~200 MB
requires_root
false
true
n/a
true
self_hosted_only
false
false
true
true
config_files
0
YAML
web ui
extensive
starting_price_usd
0
15/host
0 (self)
1,750/yr
m
Matt <admin@netwatchlabs.com>
commit 8c3f1a2 · solo founder · signed
Built by Matt — a solo founder who got tired of configuring Nagios and paying $15/host/month for Datadog just to check if his servers can reach the internet.
NetWatch is built with Rust (agent + API) and Next.js (dashboard). The agent binary is ~5 MB and uses zero dependencies at runtime.
// 05 · ready when you are
signup, install, monitor —
all free.
$
curl -sSL get.netwatch.cloud | sh