What Network Tool does.

Everything your team needs to watch the network and fix what breaks, in one application on your own server.

Overview

One screen that says whether everything is fine, and if not, where to look.

One look tells you how many problems are open and which devices they are on.

One coloured square per device: reachable, problem, unreachable.
Current traffic, average latency, flows, syslog and traps in the last minute.
Live over a WebSocket that pushes one update per second.
Probe flow: every poll drawn as an envelope that arrives, stutters or hits a closed port.

Monitoring

Every device, service and web check is polled on its own schedule; whatever stops answering becomes a problem on screen.

Devices list with name, address, type, group, latency and state for each monitored device.

Problems

Every open problem, with the real cause on top and what it explains underneath.

Zabbix-style severities; acknowledge without closing.
Problems behind a down parent device are listed as suppressed.
Event history kept in PostgreSQL.

Service tree

Business services built from devices, web checks, voice probes and other services.

Rules: all required, any one is enough, or at least N.
24 h / 7 d / 30 d availability against an optional SLA target.
A broken service shows which member broke it.

Maintenance windows

Planned work that should not wake anyone up.

Once or repeating: daily or on chosen weekdays; local time kept across daylight saving.
Problems are listed but not counted, mailed or acted on; the outage is left out of availability.
“Start maintenance” on a device page opens a one-off window right away.

Devices

Every monitored device with its latency, state, template and type.

ICMP, SNMP v1/v2c/v3 and TCP service checks per device.
Device type detected from SNMP, ports, MAC prefix and name.
New devices seen on the network wait for approval, or are added automatically.
Device page: charts, switch port panel, LLDP/CDP neighbours, hardware, SSH and config.

Agent machines

Kubernetes nodes and servers that run the Network Tool agent.

CPU, memory, load, PSI, temperature, disks, NICs, TCP sockets and retransmits.
29 node rules evaluated by the collector.
Editable chart panels over any metric and time range.

Virtualisation

VMware vCenter 6.5+
VMware ESXi 7+
Proxmox VE

VMware vSphere and Proxmox VE hosts, virtual machines and datastores.

vCenter 6.5+ or a single ESXi 7+, and Proxmox VE over REST.
Datastore 90 % / 95 % full and host disconnected become problems.
Hyper-V is not covered.

Web & certificates

Is the page up, is it fast, and when does its certificate expire.

HTTP status, expected text and response time.
Two failures in a row: down. Three slow answers: slow.
TLS expiry warning, then high at 7 days, disaster once expired.

IP SLA / VoIP

Call quality towards a voice gateway, measured from the server.

Jitter, loss and MOS (G.107 E-model).
Two bad measurements in a row raise a problem.
Cisco IP SLA results read over SNMP.

IP addresses

Which addresses are used, by whom, and on which switch port.

Subnet usage and DHCP scopes.
MAC to switch port from the switches’ bridge tables.
Built from monitored devices, ARP, scans and new devices.

Hardware inventory

Serial numbers, models, software versions and warranty dates.

ENTITY-MIB chassis, module, power supply and fan serials, read every six hours.
Warranty and support end warn 90 days ahead.
CSV export.

Maps

Your network and the software running on it, drawn from live data rather than a diagram someone has to keep up to date.

Software map: machines arranged in layers by role, with the services they run and the connections between them.

Network map

A 3D map in three tiers: the Internet, network devices and the local network.

Device shapes by type, status rings, link thickness by traffic.
Conversations drawn both ways; failing HTTP answers in red.
Cables between devices from LLDP/CDP neighbour tables.

Software map

Which machine runs which service on which port, and who talks to whom.

Layers by role: clients, API/web, queues, databases, servers, Kubernetes.
A catalogue of about sixty technologies recognised from their ports.
Draw by hand and save a design next to the live view.

Traffic

See who uses the bandwidth, over what and how much: live on the server, and from your routers’ flow records.

Live traffic

The traffic passing the server right now.

Packet capture with libpcap: top sources and destinations, conversations, protocols, ports.
Hour-of-week baselines raise a problem when an address goes far above its usual.

NetFlow flows

Traffic as your routers and switches report it.

NetFlow v5 / v9 / IPFIX on udp/2055, sFlow v5 on udp/6343.
Exporters, top talkers, recent flows; sFlow counters fill an interface table.

HTTP requests

Plain HTTP requests and answers, read from the wire.

Method, URL, headers, body, status and duration.
TCP reassembly, chunked bodies, gzip/deflate.
Plain HTTP only: TLS content cannot be read.

Interfaces

Every SNMP interface with its speed, rate, utilisation and errors.

64-bit counters; rates from consecutive samples.
Interface down and utilisation thresholds raise problems.

Logs

Syslog, SNMP traps and firewall logs from every device, kept in one place and searchable.

Firewall page: allowed and denied counts, an hourly chart, top denied sources and ports.

Syslog

Log lines from every device in one list.

udp/514: RFC 3164, RFC 5424 and Cisco IOS formats.
Critical lines become events; kept 90 days by default.

SNMP traps

Traps and informs as devices send them.

udp/162, v1 and v2c, well-known trap names.
Interface name taken from the trap.

Firewall

Rule usage and denied traffic from your firewall’s logs.

FortiGate, Cisco ASA/FTD, pfSense, iptables/nftables, MikroTik.
Most-denied sources, destinations and ports; hourly histogram.
Denies to 20+ ports in 5 minutes are flagged as a scan.

Tools

From an address you do not recognise to the cause: one page per address, diagnostics, scans, config backups and reports.

Reports page: availability per device, problem counts, capacity forecast and traffic over the period.

Machine 360

Everything known about one address on one page.

Identity, switch port and MAC, open problems.
One timeline from eight sources; sources with nothing to say are named too.

Diagnostics

Ping, traceroute, DNS and port checks without opening a terminal.

Stoppable jobs; recent runs can be reopened.
Traceroute with reverse DNS per hop.

Scan network

Find what is on a subnet and add it with one click.

ARP table and ICMP sweep, with an optional TCP port probe.
Compare two scans: new, gone and changed devices, opened and closed ports.

Build network algorithm

Draw a network design in 3D and check it against the rules.

LAN and Internet zones, gateways, switches, servers and clients by drag and drop.
Checks CIDR overlap, addresses inside their subnet, a gateway per LAN and paths to the Internet.

Config backups

Device configurations saved over SSH, with what changed.

Cisco IOS/NX-OS, MikroTik, Junos, Huawei, Linux or custom commands.
Nightly and on demand; a change raises an event and can send mail.
Compliance rules such as “telnet must be off”.

Reports

Availability, problems and traffic for a period, ready to send.

24 h / 7 d / 30 d availability per device, with an SLA target.
CSV export and print to PDF; scheduled reports by e-mail.
Capacity forecast for disks and memory.

Administration

Who can see and change what, how alerts reach people, and a record of every change.

Users and roles: users with their roles, and the role matrix of areas with None, View or Manage rights.

Management

Every collector switched on or off from the screen, with its interval and thresholds.

Changes apply at once and are saved in PostgreSQL.
Company name, subtitle and logo for screens and e-mail.

E-mail alerts

Who gets told, about what, and how often.

Built-in SMTP client; WhatsApp through the Meta Cloud API; webhooks for Slack, Teams and ticketing.
Wait time, repeat interval, resolved notice and heartbeat rules.
Scheduled daily, weekly or monthly reports.