Solution · Bandwidth and traffic

How full the link is, and who filled it


Bandwidth monitoring is two questions wearing one name. The counter answers the first. The flow records answer the second. SysPeep keeps both on the same device page.


Key points

  • SNMP interface scans find every port and keep per-interface trend history for the whole switching and routing estate.
  • Two predefined sensors turn interface use in and out into a threshold in percent of the link speed the device reports, seeded with a warning at 80%.
  • NetFlow, sFlow and IPFIX show top talkers, conversations and QoS classes for the same interfaces, included in every edition.
  • Heat strips by hour and a seven-by-24 pattern of the last four weeks show the surge and the backup window as a pattern, not a one-off.
  • A capacity report with countdowns shown as dates, and a gap in the data drawn as a gap, never as quiet.

On this page: The counter side The flow side Patterns, not incidents WAN and branches Capacity Learn the subject


"The network is slow" is the complaint; "the uplink is at 94% and it is the backup server talking to the file server" is the diagnosis. Getting from one to the other needs the interface counters for the first half and the flow records for the second, and it needs them in one place, because the person asking will not wait while you open a second console.


The counter side: every interface, named, with history


What SysPeep doesWhat it means on call
SNMP v2c and v3 interface scans on switches, routers, firewalls and VPN concentratorsUtilization and error rates for every port, with per-interface trend history, no vendor-specific tooling required
Utilization sensors in percent of link speed, in and out, seeded with a warning at 80%A saturating uplink alerts like anything else without a bandwidth figure typed in per interface
Per-instance sensors: one sensor for a set of interfaces whose status is the worst instance's, each instance its own chart lineOne saturated port among forty-eight is reported as the problem, not averaged into a calm total
Adding interfaces adds nothing to the license: licensing counts devices, and sensors are unlimitedThe thirtieth interface on a switch you already monitor is free, so nobody rations the ports they watch

The flow side: who, to whom, how much


NetFlow v5 and v9, IPFIX and sFlow v5 are collected by the same service, with listeners on the server and collection at remote probes. The Flows page shows top talkers, conversations, interfaces and QoS (DSCP) classes, filtered by site, time range and exporter, with interfaces named as the device's SNMP sensors name them and utilization shown against the speed the device reports. The NetFlow analyzer page covers it in full.


The two halves meet on the device page: the interface sensor raises the alert, and the flow chip beside it links to the traffic that explains it. A router whose export stops raises a flow-silence event within minutes, so the explanation does not quietly go missing.

Flow analysis, included

  • NetFlow v5/v9, IPFIX, sFlow v5
  • Top talkers and conversations
  • QoS classes and traffic by country
  • In every edition, never per interface
  • Collected at branches by probes (paid editions)
"What is eating the bandwidth" answered with evidence, from the same screen that told you the bandwidth was being eaten.

Patterns, not incidents


Heat strips show interface use in and out as a percent of the link speed, one cell per hour over a day or a week and coarser over 30 and 90 days. A seven-by-24 pattern of the last four weeks shows the Monday-morning surge and the nightly backup window as what they are: a pattern. An hour the server did not observe is hatched rather than drawn as quiet, so a gap in monitoring never masquerades as a quiet network. Long ranges read hourly totals kept in their own tables, so a 90-day traffic view is cheap and never drawn from data that was already pruned.


WAN links, branches and the ISP


A QoS packet train between probes measures loss, jitter and latency across a real path, which proves whether the WAN or the VoIP complaint is the network's fault, with numbers. Remote probes (paid editions) poll a branch from the inside, collect its flows locally and report back over mutual TLS, so no inbound firewall rule is needed at the branch. When the ISP says the circuit is fine, the evidence page explains what SysPeep's reports let you put on the table.


Capacity: when to upgrade, as a date


The capacity report sits beside the scorecard, health, SLA, alerts, inventory and traffic reports, exportable as CSV in every edition and as PDF in the paid editions. Countdowns are shown as projected dates rather than as a number to convert, and every figure is bounded by the coverage that was actually measured. The library's capacity planning guides cover how to turn history into an upgrade decision.


Learn the subject


The bandwidth and utilization guides explain how full a link really is, who is using it and when to upgrade; the core network metrics guides cover bandwidth, latency, jitter and packet loss and what "good" looks like; the flow monitoring guides cover NetFlow, sFlow and IPFIX from the ground up.


Questions administrators ask about bandwidth monitoring


Does it monitor Ethernet traffic on a switch port, and can it work as an Ethernet traffic analyzer?

Yes to the first and mostly to the second. Every port on a monitored switch gets an interface sensor over SNMP with bytes in and out, utilization in percent of link speed, and error and discard counters, which is what an Ethernet traffic monitor measures. For the analyzer part, who was talking to whom on that port, the device's NetFlow, sFlow or IPFIX export fills the Flows page. What it does not do is capture packets; for payload-level analysis you need a packet capture tool and a mirror port.


Is this a net traffic monitor or net speed monitor for Windows 10 and Windows 11?

SysPeep installs on a Windows server or desktop, including Windows 10 and Windows 11, and monitors the network from there: the switches, routers, firewalls and the Windows hosts themselves over WMI. It is not a taskbar speed meter for one PC; it is the traffic monitor for the whole network, run from one Windows machine, with the free edition covering up to 5 devices.


How do I check bandwidth usage per device, per IP address and per conversation?

Per interface comes from the SNMP counters. Per device, per IP and per conversation come from flow records: enable NetFlow, sFlow or IPFIX export on the router or core switch and the Flows page shows top talkers, conversations and interfaces for any time range. A bandwidth usage tracker that only sees the interface counter can tell you the link is full; the flows tell you who filled it.


Can it monitor a router's bandwidth, or do I need a router with bandwidth monitoring built in?

Any router that answers SNMP, which is nearly all of them, gives SysPeep its per-interface traffic without any monitoring feature of its own. Routers with bandwidth monitoring built in usually mean routers that export flow records, and if yours does, point the export at SysPeep and the per-host picture appears as well. Router bandwidth management, meaning shaping and rate limits, stays on the router; SysPeep shows you whether the limits are doing what you intended.


How is bandwidth utilization calculated? The formula SysPeep uses

Bytes counted on the interface during the polling interval, multiplied by eight to get bits, divided by the interval in seconds to get bits per second, divided by the link speed the device reports, as a percentage. In and out are calculated separately, because a full-duplex link can be saturated in one direction and idle in the other. The two predefined utilization sensors come seeded with a warning at 80%.


What happens when a link has exceeded its bandwidth limits?

The utilization sensor crosses its threshold and raises an alert on the device, with debounce so a five-second burst pages nobody. The alert names the interface by the name the device gives it, and the flow chip beside it links to the top talkers for the same minutes, so the first thing you see is the link and the second thing is the cause.


Is SysPeep bandwidth management software or a bandwidth manager?

No. A bandwidth manager shapes, limits or prioritises traffic, and that job belongs on the router or firewall. SysPeep measures: how much each link carries, who is using it, and whether the QoS classes you configured are getting the share you meant them to get. Measure first; most bandwidth management decisions turn out differently once the Flows page has run for a week.


How can I increase bandwidth, or find out whether I actually need to?

Look at sustained peak utilization, not the average. A link that averages 20% but sits at 95% every morning from nine to ten needs either more bandwidth or a schedule change, and the seven-by-24 traffic pattern shows which. Before buying a bigger circuit, check the top talkers for the hour: a backup job moved to 2 a.m. is cheaper than a WAN upgrade.


Is there a real time bandwidth monitor download, and a NetFlow traffic analyser (analyzer) download?

One download, the SysPeep installer for Windows, contains both. The interface sensors give the real-time bandwidth view on a 30-second clock; the NetFlow, sFlow and IPFIX collector is part of the same service and is included in every edition, the free one too. There is no separate analyzer to download or license.


Does it work as a LAN traffic monitor and a WAN bandwidth monitor at the same time?

Yes. LAN ports on the access and core switches, WAN interfaces on the routers and firewalls, and high-bandwidth connections between sites are all interfaces with counters, and the Flows page can be filtered by site and by exporter. Between sites, a QoS packet train between two probes measures loss, jitter and latency across the real WAN path, which is what proves whether the circuit or the application is at fault.


Can it monitor internet usage and network data usage?

Internet usage across the organisation is the traffic on the WAN interface, which the utilization sensor and the Flows page cover, including the countries the busiest senders are placed in from a bundled offline database. Network data usage per host comes from the flow records. It does not monitor individual users' browsing, and it is not an employee monitoring product.


Does it monitor Wi-Fi bandwidth and wireless traffic?

Access points and wireless controllers that answer SNMP expose their radio interfaces and client counts like any other device, so Wi-Fi bandwidth is monitored the same way as a wired port. For who is using the wireless bandwidth, the flows have to come from the switch or router the access points uplink to, which is where the traffic becomes visible.


Is there a bandwidth meter that shows the current rate on an interface as a number?

The interface sensor's current value is the rate in bits per second over the last polling interval, shown in the console with the same unit model everywhere, so 9.7 Mbps reads the same on the device page, the dashboard and the notification. The graph beside it shows the history at whatever range you pick.


Can bandwidth monitoring be free?

On up to 5 devices, yes, at full depth: every interface, flow analysis included, on a license that never expires. The Free edition page lists exactly what is and is not included.


What about bandwidth monitoring on Linux servers?

A Linux host's interfaces are read over SNMP if its agent is enabled, or over SSH, where whatever a command prints becomes the reading. Either way the host is one monitored device with unlimited sensors, and its traffic appears beside the switches it connects to.


Does it monitor network traffic by application or by port?

By port, protocol, host, conversation and QoS (DSCP) class, from the flow records. It does not perform deep packet inspection or application signatures, so "traffic by application" means traffic by the ports and classes the application uses, which is enough for most capacity questions and honest about the rest.


Is there a bandwidth usage report, and can it be scheduled?

The traffic report sits beside the scorecard, health, SLA, alerts, capacity and inventory reports. Every edition exports it as CSV; the paid editions export PDF and deliver it on a schedule by email. Every total says how much of the window was observed.


Does it keep bandwidth usage history for 30 or 90 days?

Long ranges read hourly totals kept in their own tables with their own retention, so a 30-day or 90-day traffic view costs hourly rows rather than per-minute records. Flow history has its own retention setting and a storage ceiling; metric history keeps 90 days by default. The Free edition's console shows the last 30 days.


Is it a bandwidth consumption monitor and a bandwidth usage tracker as well as a meter?

Yes. The interface sensors track consumption over time, the Flows page tracks usage by host and conversation, and both keep history, so the question "how much did we use last month and who used it" has an answer with a graph.


Can it monitor outgoing network traffic and firewall traffic on Windows 10, Windows 11 and Windows Server 2012 R2?

A Windows host's own interface counters, in and out, are read over WMI on any supported Windows version, so outgoing traffic from a specific machine is a sensor. Firewall traffic, meaning what the firewall passed and dropped, comes from the firewall itself over SNMP and from its syslog lines, not from the workstation.


Does it monitor high bandwidth connections, internal network bandwidth and SD-WAN links?

A 10-gigabit or 100-gigabit interface is an interface with a bigger speed value; utilization is still a percentage of what the device reports. Internal network bandwidth between switches is the uplink counters; SD-WAN links are monitored on the edge device's interfaces and tunnels, with a QoS packet train between probes measuring the overlay path.


How is the bandwidth of a network measured, and is there an SNMP traffic grapher in it?

By reading each interface's byte counters over SNMP at a fixed interval and dividing by the time and the link speed, which is what the classic SNMP traffic grapher did and what the interface sensors do on a 30-second clock, with the graph beside the number. Flow records add who used it; a packet train between probes adds what the path can actually carry.


Scan one switch and watch the uplink graph fill in

Download SysPeep, let discovery find the switch, and the interface sensors come with their 80% warning already set. Up to 5 devices are free at full depth, flow analysis included; the 30-day trial unlocks every feature except clustering.

Download Free TrialThe NetFlow analyzer