CCNA 200-301

CCNP Enterprise

CCNP Security

CCIE Enterprise Lab

CCIE Security Lab

CCNP Service Provider

CCNP Data Center

CCNP Collaboration

CCIE DC Lab

Measuring network latency directly from a Cisco router allows network engineers to isolate WAN slowdowns, verify service provider SLAs, and distinguish between transit link congestion and local device bottlenecks. Whether you need an immediate round-trip time (RTT) measurement between two endpoints or continuous telemetry tracking microsecond jitter across voice and video paths, Cisco IOS and IOS-XE provide built-in diagnostic tools for every stage of analysis.

Checking Real-Time Latency via Cisco CLI Commands

When you need an immediate, ad-hoc measurement of round-trip time (RTT) from a Cisco router to a remote host or next-hop gateway, the command-line interface (CLI) offers three sequential diagnostic steps.

Step 1: Run a Standard ICMP Ping for Quick RTT Baseline

From user EXEC (>) or privileged EXEC (#) mode, run the standard ping command to send five 100-byte ICMP Echo Request packets to the target IP address:

Router# ping 8.8.8.8
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 8.8.8.8, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 12/14/18 ms

Evaluate the output using two indicators:
* Response Characters: Each ! indicates a successful reply received within the 2-second timeout. A . indicates a timeout (packet loss), while U denotes that an ICMP Destination Unreachable message was returned.
* RTT Summary (min/avg/max): Displayed at the bottom in milliseconds (ms), showing the fastest, average, and slowest round-trip times across the sample.

Step 2: Use Extended Ping to Test Specific Paths and Payloads

A standard ping uses the outbound interface IP chosen by the routing table and sends only five small packets, which often misses intermittent latency spikes or MTU-related delays. Run an extended ping in privileged EXEC mode either interactively (by typing ping and pressing Enter) or via a single inline command:

Router# ping 10.20.30.1 source GigabitEthernet0/0/1 repeat 100 size 1400 df-bit
  • source <interface | ip>: Forces the router to originate probes from a specific LAN or tunnel interface, verifying latency across policy-routed paths or IPsec VPNs.
  • repeat <count>: Increases the sample size (for example, 100 or 500 packets) to expose periodic latency jitter that a 5-packet test misses.
  • size <bytes> and df-bit: Tests whether larger packets trigger serialization delays or fragmentation issues along the path. If the destination resides inside a specific VRF, insert vrf <vrf-name> immediately after ping.

Step 3: Isolate Hop-by-Hop Latency with Traceroute

If ping reveals elevated end-to-end RTT, use traceroute to identify exactly which intermediate router or carrier link introduces the delay:

Router# traceroute 10.20.30.1 source GigabitEthernet0/0/1 numeric
Type escape sequence to abort.
Tracing the route to 10.20.30.1
VRF info: (vrf in name/id, vrf out name/id)
  1 192.168.1.2 1 msec 1 msec 1 msec
  2 172.16.10.1 4 msec 3 msec 4 msec
  3 10.20.30.1 48 msec 47 msec 49 msec

Cisco IOS traceroute transmits three UDP probes per hop with incrementing Time-to-Live (TTL) values and prints the RTT in msec for each probe. Including the numeric keyword skips reverse DNS lookups on each hop, ensuring the CLI outputs probe results immediately without DNS timeout delays. Looking for a sudden, sustained jump in RTT from one hop to the next pinpoints the exact segment responsible for the latency increase.


Understanding show interfaces Delay (DLY) vs. Actual Network Latency

After running CLI ping tests, administrators frequently inspect interface counters using show interfaces and notice a DLY metric in the output. Because this value is often confused with live link latency, it is important to clarify how Cisco IOS treats interface delay parameters.

Does show interfaces display real-time network latency in Cisco IOS?

No. When you execute show interfaces <interface>, the DLY value shown on the third line (for example, MTU 1500 bytes, BW 1000000 Kbit/sec, DLY 10 usec) is a static, hardcoded constant measured in microseconds (usec). It never fluctuates based on live traffic load, physical propagation time, or queuing delay.

What does the DLY metric in show interfaces actually represent?

The DLY field represents a default theoretical media delay assigned according to the interface type and bandwidth—such as 10 usec for GigabitEthernet, 100 usec for FastEthernet, and 20000 usec for a Serial T1 link. Its purpose in Cisco IOS is to serve as a routing metric component for EIGRP (Enhanced Interior Gateway Routing Protocol), which calculates composite path costs using the cumulative interface delay along a route alongside minimum bandwidth.

Does changing the delay command on an interface affect actual packet latency?

Not directly on that physical link. Modifying the parameter with the interface subcommand delay <tens-of-microseconds> does not alter hardware serialization speed or physical link latency. However, changing the delay value alters EIGRP metric calculations, which can shift traffic onto an alternate network path that has different real-world latency characteristics.


Configuring Cisco IP SLA for Continuous Latency and Jitter Monitoring

While ad-hoc ping and traceroute commands measure latency at a single moment, tracking trends over time requires automated probing. You can configure Cisco IP SLA directly on the router using the templates below to continuously measure round-trip latency, one-way delay, and jitter.

Template 1: ICMP Echo Operation (Continuous RTT Latency Tracking)

Use this template to monitor round-trip latency to any IP target (even non-Cisco devices or public DNS resolvers) without requiring special software on the destination host:

configure terminal
! Define IP SLA operation 10
ip sla 10
 icmp-echo <DESTINATION_IP> source-ip <SOURCE_IP_OR_INTERFACE>
 frequency <INTERVAL_IN_SECONDS_E.G._60>
 timeout 5000
 threshold <RTT_ALERT_THRESHOLD_MS_E.G._100>
 history hours-of-statistics-kept 24
 history distributions-of-statistics-kept 5
exit
! Schedule operation 10 to run continuously starting immediately
ip sla schedule 10 life forever start-time now
end
write memory

Template 2: UDP Jitter Operation (RTT, One-Way Latency, Jitter, and Packet Loss)

Use this template when monitoring VoIP or real-time video links. It requires enabling ip sla responder on the remote Cisco router so both endpoints can timestamp packets for one-way latency and positive/negative jitter calculations:

! --- Run on Destination (Responder) Cisco Router ---
configure terminal
ip sla responder
end

! --- Run on Source Cisco Router ---
configure terminal
ip sla 20
 udp-jitter <RESPONDER_IP> <UDP_PORT_E.G._16384> source-ip <SOURCE_IP> num-packets 20 interval 20
 tos 184
 frequency 30
 threshold 150
exit
ip sla schedule 20 life forever start-time now
end
write memory

CLI Commands to View IP SLA Latency Statistics

Run the following privileged EXEC commands to inspect the latest RTT completion time, min/avg/max jitter, and historical buckets:

! View latest RTT result and return code for ICMP Echo probe
show ip sla statistics 10

! View detailed jitter, one-way latency, and packet loss metrics
show ip sla statistics 20 details

! View historical latency tables collected over time
show ip sla history 10

How to Troubleshoot High Latency and Packet Loss on a Cisco Router

When your CLI tests or IP SLA probes confirm high latency or packet drops across a Cisco router, follow this four-step diagnostic sequence to isolate whether the bottleneck is caused by the control plane, physical layer errors, interface congestion, or QoS policies.

Step 1: Differentiate Control-Plane Latency from Transit Data-Plane Latency

Pings sent to one of the router’s own IP addresses are punted to the control-plane CPU and may be rate-limited by Control Plane Policing (CoPP), whereas traffic passing through the router is forwarded in hardware via Cisco Express Forwarding (CEF). Check whether high CPU utilization is delaying control-plane responses or software-switched packets:

Router# show processes cpu sorted 5sec
CPU utilization for five seconds: 84%/12%; one minute: 79%; five minutes: 75%
 PID Runtime(ms)     Invoked      uSecs   5Sec   1Min   5Min TTY Process
  89     4829100    12049102        400 42.15% 38.40% 36.12%   0 IP Input

On Cisco IOS-XE platforms (such as ASR 1000, ISR 4000, and Catalyst 8000 Series), also check data-plane Quantum Flow Processor (QFP) load using show platform hardware qfp active datapath utilization to verify that hardware forwarding engines are not saturated.

Step 2: Check for Physical Layer Errors and Duplex Mismatches

Damaged cables, failing optics, or speed/duplex mismatches force layer-2 frames to be dropped and retransmitted by upper-layer protocols, manifesting as erratic latency spikes. Inspect the error counters on the ingress and egress interfaces:

Router# show interfaces GigabitEthernet0/0/0 | include duplex|errors|CRC|drops
  Full-duplex, 1000Mb/s, link type is auto, media type is RJ45
  0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
  Input queue: 0/375/0/0 (size/max/drops/flushes); Total output drops: 14820

If input errors, CRC, frame, or runts are incrementing, resolve the physical cabling, SFP transceiver, or duplex mismatch before investigating higher-layer configurations.

Step 3: Diagnose Interface Output Queue Drops and Buffer Congestion

Notice the Total output drops counter in the show interfaces output above. When bursty traffic exceeds the egress bandwidth of a WAN link, packets wait in the interface output buffer—directly increasing queuing latency (bufferbloat)—until the queue fills and tail drops occur. On Cisco IOS-XE routers, inspect the exact hardware drop causes across the forwarding plane:

Router# show platform hardware qfp active statistics drop
-------------------------------------------------------------------------
Global Drop Stats                         Packets                  Octets
-------------------------------------------------------------------------
TailDrop                                    14820                19414200
ipPolicerDrop                                3410                 4126100

High TailDrop counts confirm that egress bandwidth saturation is delaying and dropping packets during traffic bursts.

Step 4: Audit Active QoS Shaping and Policing Policies

Traffic shapers hold excess packets in memory queues to smooth out bursts to a committed information rate (CIR), intentionally trading higher latency for fewer carrier drops. Conversely, policers immediately drop or remark excess traffic. Inspect active Quality of Service (QoS) policies on the egress interface:

Router# show policy-map interface GigabitEthernet0/0/0

Check each traffic class for queue limit, (pkts output/bytes output), (depth/total drops/no-buffer drops), and policer exceeded or violated counters. If latency spikes correlate with rising queue depth inside a shaped class, increase the shaped bandwidth rate (if the WAN circuit allows it) or prioritize latency-sensitive traffic using a Low Latency Queuing (priority) class.


Best Network Monitoring Tools for Cisco Router Latency and IP SLA

While CLI commands are essential for live troubleshooting, enterprise networks rely on external Network Performance Monitoring (NPM) software to poll Cisco routers via SNMP, NetFlow, and IP SLA MIBs for long-term visualization and alerting:

  1. SolarWinds Network Performance Monitor (NPM) & VoIP and Network Quality Manager (VNQM): Polls Cisco routers via SNMP to graph baseline ICMP response time, visualizes hop-by-hop internet latency using NetPath, and integrates directly with Cisco IP SLA (CISCO-RTTMON-MIB) to automate WAN jitter, latency, and MOS score alerting.
  2. Paessler PRTG Network Monitor: Includes dedicated Cisco IP SLA SNMP sensors alongside high-frequency Ping and System Health sensors, allowing administrators to plot one-way latency, jitter, and packet loss per SLA operation ID on real-time dashboards.
  3. Cisco Catalyst Center & Cisco ThousandEyes: Combines native IOS-XE streaming telemetry with active synthetic monitoring. Supported Cisco ISR 4000, ASR 1000, and Catalyst 8000 Series routers can host embedded ThousandEyes Enterprise Agents directly in a container on the router to visualize end-to-end BGP, WAN, and SaaS application latency.
  4. ManageEngine OpManager: Automatically discovers pre-configured IP SLA probes on Cisco routers via SNMP and generates historical WAN RTT trend reports, threshold violation alerts, and voice quality monitors without manual OID mapping.
  5. Zabbix and LibreNMS (Open-Source Platforms): Free, self-hosted monitoring platforms that ship with built-in Cisco IOS/IOS-XE discovery templates. Both automatically index CISCO-RTTMON-MIB operations and correlate RTT latency spikes with router CPU utilization and interface queue drops on unified graphs.

Official Cisco IOS Command References and Documentation

For syntax specifications, platform-specific release constraints, and SNMP schema definitions across Cisco IOS (15.x) and Cisco IOS-XE (16.x / 17.x), consult the following authoritative Cisco documentation resources:

  • Cisco IOS IP SLAs Configuration Guide: Defines the architecture and step-by-step configuration procedures for icmp-echo, udp-jitter, icmp-jitter, and tcp-connect operations, IP SLA responder setup, proactive threshold monitoring (ip sla reaction-configuration), and Enhanced Object Tracking integration (track <object-number> ip sla <operation-number>).
  • Cisco IOS IP SLAs Command Reference: Provides the complete alphabetical command syntax, parameter ranges, default values, and command mode hierarchies for all ip sla configuration commands and show ip sla diagnostic commands.
  • Cisco IOS Configuration Fundamentals Command Reference (ping and traceroute): Documents standard and extended EXEC command flags, VRF-aware syntax (ping vrf, traceroute vrf), protocol-specific options, and complete tables of CLI response status characters (!, ., U, Q, M, ?, &).
  • Cisco SNMP Object Navigator (CISCO-RTTMON-MIB): Details the official ASN.1 MIB structure (OID 1.3.6.1.4.1.9.9.42) used to query rttMonLatestRttOperCompletionTime, rttMonLatestJitterOperAvgJitter, and operation return codes from external SNMP managers.

Conclusion

Checking latency on a Cisco router starts with quick CLI validation using extended ping and traceroute—while avoiding the misconception that the static DLY metric in show interfaces reflects live network delay. By pairing continuous on-box IP SLA probes with structured control-plane and interface queue diagnostics, and feeding those metrics into an SNMP-aware monitoring platform, you can pinpoint network slowdowns and maintain predictable WAN performance.

Please follow and like us:
Last modified: November 17, 2026

Author

Comments

Write a Reply or Comment

Your email address will not be published.