Features

Everything you need to monitor your fleet

Real-time metrics, container control, uptime checks, alerts, groups, and organizations with SAML single sign-on - all in one dashboard.

Core monitoring

CPU, RAM, disk and network default to one-second collection; GPU defaults to five seconds. Collection intervals are configurable with statsd. Stored history resolution and retention depend on your plan, with average and P95 series, live mode, and CSV export.

  • CPU per core with load 1 / 5 / 15 and frequency, where supported by the platform
  • GPU metrics depend on hardware and OS. NVIDIA requires nvidia-smi; AMD and Intel live metrics are unavailable on Windows
  • Online / Connected / Offline states with last-seen timestamps
  • History windows up to 365d on Organization, with average and P95, plus live mode
  • CSV export on every chart; Free keeps 14 days at 60 s resolution, Pro 180 days at 1 s, Organization 365 days at 1 s

CPU

42%

GPU

18%

RAM

68%

Disk

31%

CPU history
5m1h24h7d30d365d
Avg: 34%P95: 72%● liveCSV ↓

Machines

Rows, grid, or cards - drag to reorder, nest into groups, and drill into every machine with six dedicated tabs.

  • Per-core CPU, memory pressure and swap, disk per partition with read / write rates, ZFS pools
  • Hardware inventory on the Device tab: P / E cores, GPU VRAM, DIMM type and speed, disks, NICs with link speed
  • Per-machine tabs: Dashboard · History · Containers · Control · Device · Management
  • Inline rename, group assignment, unregister, forget history, delete
RowsGridCards

prod-server-01

Ubuntu 22.04

42%18%68%31%
now

Mac Mini M2

macOS 15.3

12%8%54%47%
now

dev-workstation

Windows 11

78%62%85%62%
1m ago

backup-node

Debian 12

0%0%0%0%
3h ago

Docker & containers

See every container's CPU, memory, health, and status across every host, collected every 3 seconds by default. Start, stop, restart, pause, pull images, remove, and stream logs - from any device.

  • Containers grouped by machine, nested under group hierarchy
  • Start / Stop / Restart / Pause / Unpause / Pull / Remove per container
  • Details popup with state, image, ports, health, and IO rates
  • Live logs viewer - no SSH or terminal needed
prod-server-018 containers · 7 running
nginx-proxyhealthy2.1%128M
postgres-15healthy8.4%512M
redis-cachehealthy0.3%64M
api-workerunhealthy42.1%1.2G
legacy-job--%-

Proxmox guests

Run statsd on a Proxmox node and its VMs and LXC containers show up beside the host. Auto-detected - no agent inside each guest.

  • VMs and LXC containers collected every 5 seconds by default
  • Detected automatically on the node - nothing to configure
  • Guests listed under their host in the same dashboard
pve-node-01Proxmox VE4 guests · 3 running
VM100ubuntu-web12%2.1G
VM101win-build48%8.0G
LXC200pihole1%96M
LXC201backup--
Refresh: 5sDetected automatically

Hosts

Watch any endpoint over HTTP, ICMP ping, or a MongoDB connection - public sites, internal services, third-party APIs. Track response times, status codes, and uptime.

  • HTTP, ICMP, and MongoDB checks, as often as every 10 seconds
  • Custom method, headers, expected status codes, body match, and timeout
  • Success and failure thresholds before a host changes state
  • Uptime bars for 24h / 7d / 30d / 90d, response-time and connectivity charts

Hosts

5 watched · 4 up · 1 down

https://api.example.com/health20084ms99.98%
icmp://10.0.0.1ping2ms100%
mongodb://db.internal:27017ok11ms99.91%
https://billing.internal/healthz503-98.12%
https://docs.example.com200201ms99.74%
Last check: 8s agoInterval: 10s

Alerts

Pro

Metric rules held for 1 to 15 minutes, status rules for machines, containers, and hosts. Every incident gets a timeline and can be resolved by hand. Supported destinations include email, Slack, Discord, Microsoft Teams, Mattermost, Rocket.Chat, BotComm, Telegram, ntfy, Pushover, Gotify and webhooks.

  • Metric rules on CPU / GPU / RAM / Disk with >, <, =, held 1 to 15 minutes
  • Status rules for machine, container, and host
  • Scope to all machines, a group, or specific ones - with exclusions
  • Incidents with a timeline and manual resolve
  • Flood control: 30-minute dedup, at most 20 notifications an hour per channel, then an hourly digest

New alert rule

High CPU usage
CPU %
>
80
for
5m
Group: Production 1 excluded

Recent incidents

prod-server-0110:42:03resolve
dev-workstationYesterday 18:22Yesterday 18:35
backup-nodeNov 14 09:11Nov 14 09:14

Groups

Nest groups to mirror your infrastructure - regions, environments, organizations. Drag to reorder, and deleting a group reparents its contents.

  • Parent / child hierarchy, nested as deep as you need
  • Drag-reorder on Dashboard and Machines pages
  • Group cards show online / health ratios at a glance

Groups

Production88/8
US-East55/5
web33/3
db22/2
EU-West33/3
Staging43/4
Development21/2

Organizations and sharing

Every account starts with a personal organization. Share machines and checks with people you invite, using Owner, Admin and Member roles. Pro includes three seats; the Organization plan adds more organizations and SAML single sign-on.

  • Share individual machines and hosts with an organization
  • Owners send email invitations. Recipients sign in with the invited email and accept before gaining access
  • Members use shared resources without managing the organization; owners and admins manage access according to their roles
  • Pro includes 3 seats - you and two more. Organization has no per-seat charge, with up to 200 members in each organization
  • SAML on Organization: connect your identity provider, verify email domains, test sign-in and enable the connection
  • Require invitations or automatically add SAML users as Members after sign-in and any required two-factor authentication

Organization · Infra

3 members

Yyouowner
Aalexadmin
Ssammember

Pending invitation

jordan@example.commember

Access starts after acceptance

Shared with this organization

Machines

6

Hosts

4

Single sign-on · SAML

Active
example.comVerified
MembershipInvitation required

Two-factor and sessions

TOTP two-factor on every plan. Agents link with a device code, so your password never goes on a machine, and changing it signs every other session out.

  • TOTP with QR setup and 10 recovery codes, on all plans
  • Sign in with a password or a one-time code emailed to you; two-factor still applies
  • Agents authenticate with a device code - statsd login prints a link, you approve it
  • Refresh tokens rotate; every session is revoked on password change

Two-factor authentication

Scan with your authenticator

TOTP · 6 digits · 30s

482 913
Recovery codes: 10 unusedSessions: 3 active

Dashboard

One overview with fleet health, a ranked list of what needs attention, clickable summary cards, active incidents, group cards, and the last day of incidents.

  • Needs attention: disks filling up with a days-to-full forecast, unhealthy containers, memory pressure, sustained CPU, silent agents, hosts down. No rules needed, on every plan
  • Recent incidents across alerts and hosts in one feed, resolvable in place
  • System-wide health dot: green / orange / red / gray
  • Clickable cards for machines, containers, health, alerts
  • Command palette (Cmd / Ctrl K), light / dark / system theme, installable as a PWA

Dashboard

Overview of all systems

Healthy

Machines

14

13 online · 1 offline

Containers

47

42 running · 5 stopped

Health

1 unhealthy

41 healthy · 1 unhealthy

Alerts

1 active

6 configured · 1 firing

Needs attention

nas-01disk 96% · full in ~11 days96%
mac-studiomemory pressure · 6.1 GB swappressure
edge-cacheno metrics for 4 minstale

Active incidents

High CPU usageprod-server-01

Explore monitoring for your systems

Docker monitoring →

See which containers are running, how much CPU and memory they use, and whether their health checks are passing. System Stats brings Docker and machine metrics into the same dashboard.

Proxmox monitoring →

Keep Proxmox guest state and resource usage beside the machine that runs them. Install statsd on the node to discover virtual machines and LXC containers through the local Proxmox tools.

Linux server monitoring →

Monitor Linux servers, ARM machines and Raspberry Pi from one dashboard. The statsd agent runs as a background service and collects the system metrics you need to investigate resource pressure.

HTTP uptime monitoring →

Check whether an endpoint responds as expected, how long it takes, and when it becomes unavailable. Keep uptime and response-time history beside the machines and containers behind your service.

Install the agent and see it all

Free up to 5 machines forever. No credit card. Two minutes from landing here to live metrics streaming in.

Start free → Talk to us