Product
Every layer between the user and the app, measured on the device
One agent per device feeds four views, the health of your fleet and the alerts. Where a platform cannot measure something, the console says so instead of showing a zero.
Hopix NPM
Traffic and connection health, per app
Throughput, latency, retransmissions, failed connections and traffic dropped on the device, for every app and destination, minute by minute. Ranked findings say what to look at first.
- An overview with key indicators and ranked findings
- Traffic by app, service and destination
- Whether the delay is in the network or in the app
- TCP health on Windows and macOS: how long connections take to open, connections dropped by timeouts, and paths that silently drop large packets
- A map of what each app talks to, with replay
- CSV export of the tables
Time ranges from 15 minutes to 7 days, across up to 90 days of history, depending on the plan.
Protocols
Every protocol, as client and as server
A page for each protocol the device uses or serves, from DNS, HTTP and TLS to SMB, Kerberos, LDAP, RDP, SSH and databases. Each one has an overview, findings that point at the problem, its own detail, and the same charts and tables as Hopix NPM.
- More than 40 protocols recognized out of the box, split into client activity and server activity
- Findings per protocol, such as resets and failed connections on HTTPS, errors on a file share or slow responses
- Operations, error codes, versions and response times, where the system reports them
- Your own services, named once by port, with their own page in every device’s view
- Fleet servers: each server your devices use, seen from all of its clients, to tell whether it is slow for everyone or for one office
- The same pages for a group of devices
- Alerts on a protocol’s operations, error rate, timeouts and response time
Every protocol page, with its traffic, connections and findings, works on Windows, macOS and Linux. Operations, error codes and response times depend on what each system reports: today most come from Windows and some from macOS. Server activity is available on Windows and macOS.
Hopix DNS
When the name is the problem
Every lookup with its result and latency, per app and per domain, compared with the week before. In the console it is the DNS page of Protocols.
- Findings for failing domains and apps, slow resolvers and timeouts
- Flags suspicious and misconfigured lookups
- Spots domains that are new compared with the week before
- Filters by result, domain, app and latency, and CSV export
Time ranges up to 90 days, depending on the plan.
Radar
The path, hop by hop, live
Start a session from the console to the destinations you choose and watch the path draw itself: each hop with its latency, its loss and its changes.
- Live map and hop table
- Latency and loss of each hop over time
- Replay of the last 7 days
Only agent administrators can start a session, and every start is audited.
Wifi
The radio, not just the bars
Signal and link quality, link rate, band and channel, roaming between access points and every disconnection. When other devices on the same network drop at the same moment, Hopix tells you it was the access point or the router, not the laptop.
- Signal history and link rate
- Roaming and event log
- Findings for shared outages, roaming problems and weak signal
- Live radio view while you watch
- History of up to 90 days, depending on the plan
On macOS the access point ID is not available; the Linux agent has no Wi-Fi module.
Alerts
Alerts that reach the right channel
Choose from more than 50 metrics across NPM, protocols, DNS, Wi-Fi, device resources and agent status, from failed connections and a protocol’s error rate to weak Wi-Fi, a full disk or an agent offline, for one agent, a whole group, or each device of a group or of the tenant.
- Ready-made templates and a recommended starter pack
- Thresholds and ranges, a wait before firing, minimum samples
- Incidents with first, worst and last value, acknowledgment and history
- Email, Microsoft Teams, Slack and signed webhooks
- Mute for up to 7 days, re-notify until acknowledged
Groups
See a site, a team or a service at once
Group devices by hand or by rules on hostname, user and IP address, or group traffic by port. Each group gets its own NPM, Protocols and DNS views, and can drive alerts.
- Rules that match any or all of their conditions
- New devices join the groups they match on their own
- Group views add up the members, with a per-device table
Fleet management
Every agent, managed from the console
Install with one command per system, or from your RMM, GPO, Intune or Mac MDM. Then update, pause, restart, reconfigure or uninstall any agent remotely, without touching the device.
- A home page with the health of every device, the apps and destinations with problems, and the open incidents
- First steps that guide a new tenant from the first agent to the first alert
- Enrollment tokens with a use limit and an expiry
- Reinstall a device on its existing registration and keep its history, or on a new one that retires the old
- A short ID for each agent, the same in the console and on the device
- Signed automatic updates with rollback
- Settings per tenant or per agent, including the privacy profile
- CPU, memory, disk and top processes of each device
Sign-in and access
Access control your security team expects
Connect each tenant to its identity provider, require MFA, and give every person only the views they need.
- SAML 2.0 single sign-on per tenant: its users sign in only through their identity provider
- Access from your directory groups: each person gets the roles of their groups, and their account is created at first sign-in
- Passkeys and authenticator apps for local accounts, and MFA you can enforce per tenant
- Roles for tenant and agent administrators, and separate privileges for NPM, Radar and Wifi
- Invitations by one-time link: nobody chooses another person’s password
- Re-authentication for sensitive actions, and sessions you can see and close
- Audit log of sign-ins, users, agents, commands and settings, forwarded to your SIEM over syslog, encrypted by default, in JSON, CEF or LEEF
- English and Spanish, with each person’s own number format and time zone
Agents
What each agent measures
The three agents share one console and one data model. Where a system does not expose something, the console shows “Not available” rather than a zero.
| Capability | Windows | macOS | Linux |
|---|---|---|---|
| Traffic and connections (NPM) | Yes | Yes | Yes |
| DNS lookups | Yes | Yes | Yes, except encrypted DNS and DNS over TCP |
| Traffic per protocol | Yes | Yes | Yes |
| Protocol details: operations, error codes, response times | Many: file sharing, sign-in, remote desktop, web, DHCP, SQL Server and more | Some: web, TLS, QUIC, DNS and DHCP | No |
| Server activity per protocol | Yes | Yes | No |
| TCP health: connection setup, connections dropped by timeouts, large packets lost on the path | Yes; setup only for the connections the device opens | Setup and large packets lost on the path | No |
| Radar (path, hop by hop) | Yes | Yes | Yes |
| Wi-Fi | Yes | Yes, without the access point ID (hidden by macOS) | No |
| CPU, memory, disk and processes | Yes | Yes | Yes |
| Monitoring levels | All three | All three | All three; events only has extra system requirements |
| Packet contents | Never collected | Never collected | Never collected |
| Requirements | Windows 10 or 11, 64-bit | macOS 14+ on Apple silicon | x86-64, systemd: RHEL 9+, Ubuntu 22.04+, Debian 12+ |
Three monitoring levels
Full, events only, or nothing in the packet path, for each system or for a single agent. Our team sets the level you ask for, and the console shows the level each agent runs and what that level leaves out.
Light on the device
On Windows and Linux, a CPU guard pauses the most detailed measurements before they weigh on the device: by default, when they pass 5% of one core. They resume on their own.
What each agent costs, measuredSee Hopix on your own network
Book a 30-minute walkthrough. We install an agent with you and look at real data together.