Tracelet is a self-hosted appliance and a featherweight Windows agent. Watch every endpoint, network path, certificate, and machine from one executive dashboard — no cloud tenant, no data egress, no per-host metering.
Built for teams that can't ship their telemetry to someone else's cloud
Endpoints, networks, certificates and full host telemetry — correlated, alerted, and explained.
HTTP, TCP and platform health probes with response-time history, status codes and uptime SLAs at a glance.
Latency, packet loss and reachability across every target. Ping and TCP fallbacks see through Windows firewalls.
Track expiry, chains and issuers for every TLS endpoint. Get warned weeks before anything lapses — never get caught.
A 4 MB Rust agent streams CPU, memory, disk, I/O, network, top processes and event logs — securely, with near-zero overhead.
Repeat failures collapse into a single incident with accurate active-duration math. Email alerts to your own SMTP — no third party.
Ask "what's breaking and why" in plain English. Tracelet correlates failures across surfaces and platform incidents to find the root cause.
Walk in, glance once, know everything. The executive dashboard rolls up uptime, active incidents, recoveries and delivery health into an attention queue that surfaces what actually needs you.
Mint an enrollment code, hand out a signed MSI, done. The agent dedups on hardware ID so reinstalls update in place, beacons on uninstall, and a reachability probe tells offline from broken from gone.
Repeat failures collapse into one incident, then the AI assistant correlates it across surfaces and platform status to name the likely root cause. Ask in plain English; get an answer, not a wall of logs.
No cloud account. No agents calling home to a vendor. Just your appliance and your machines.
Drop the Tracelet appliance on any Windows host. It binds to loopback by default and stores everything in your own PostgreSQL.
Mint a code from Machines → Add Machine, then push the signed tracelet-agent.msi — or paste the code into the wizard. Hosts appear live.
Metrics flow in immediately. Configure thresholds, point alerts at your own SMTP, and let the AI assistant explain anything that breaks.
Tracelet was built for environments where shipping monitoring data to a SaaS vendor isn't an option. Nothing leaves your network unless you explicitly send it.
The dashboard binds to 127.0.0.1. LAN exposure is an explicit, deliberate opt-in — never the default.
Agents pin the appliance CA by file and authenticate with a per-host bearer token. No trust-store tampering.
Metrics live in your PostgreSQL. Alerts go to your SMTP. No vendor cloud, no telemetry phone-home.
Codes bind to tenant and client at mint time; only SHA-256 hashes are stored, and cross-tenant use is rejected.
No per-host metering games. One license unlocks the whole appliance.
Agentless checks for a small estate, on the house.
Unlock agent telemetry and production monitoring for one appliance.
Unlimited monitoring for growing fleets and serious operations.
Everything in Pro plus the AI assistant that explains your incidents.
msiexec). It collects CPU, memory, disk, I/O, network, top processes, event logs and host inventory with well under 1% CPU overhead.configure-remote with a routable host name; Tracelet re-mints the server certificate with the right SANs and opens the firewall for the agent channel. Already-enrolled agents keep validating — the cert change is additive.Download the appliance, enroll your first machine, and watch the metrics roll in — free for 15 endpoints, forever.