ADAM PULSE Knowledge Base
Diagnostics & troubleshooting · Latency · Carrier Responsibility

How to troubleshoot high network latency: LAN, firewall, WAN, and when it is the ISP

By Michael Obraitis, Managing Partner, and Michael Dale, CEO·Published August 12, 2026·Last reviewed August 2026·ADAM Pulse Knowledge Base

Your internet connection is online.

Bandwidth looks fine.

Nothing is technically down.

But applications feel slow.

Zoom conversations have noticeable delay.

Remote desktops hesitate.

Cloud applications take longer to respond.

The problem may be high network latency.

Latency is one of the most important network performance measurements because it describes delay.

The key troubleshooting question is not simply:

Is latency high?

It is:

> Where did the additional delay begin, and what changed?

ADAM Pulse can help USA Telecom customers establish normal network performance, preserve latency history and correlate changes with firewalls, WAN connections, carriers and application complaints.

What Is Network Latency?

Network latency is the time required for data to travel between network points.

It is commonly measured in milliseconds.

A latency measurement may represent a round trip:

Device → Destination → Device

The lower the delay, the more responsive network interactions generally feel.

What Does High Latency Feel Like?

Users may describe high latency as:

These symptoms do not automatically prove latency is responsible, but latency should be investigated.

What Causes High Network Latency?

Potential causes include:

The objective is to determine where the delay begins.

Is High Latency the Same as Slow Internet Speed?

No.

Internet speed usually refers to throughput or bandwidth.

Latency measures delay.

A connection can provide:

1 Gbps

and still have high latency.

Likewise, a lower bandwidth connection may feel responsive when it is not congested and latency is low.

Why Does Latency Matter?

Many modern applications are interactive.

Examples include:

Interactive applications depend on timely communication.

High delay can make them feel slow even when bandwidth is abundant.

What Is Good Network Latency?

There is no single number appropriate for every network path and application.

Expected latency depends on:

A better operational approach is to understand both:

What the application requires

and:

What is normal for this specific path.

Why Is a Latency Baseline Important?

Suppose a branch normally reaches a critical service in:

20 ms

Today it takes:

75 ms

A generic alert threshold of 100 ms may show no problem.

But performance has changed substantially.

A baseline reveals that change.

Can High Latency Occur Without Packet Loss?

Yes.

Traffic may arrive successfully but slowly.

This can create poor responsiveness without missing packets.

Monitor latency and packet loss separately.

Can Packet Loss Cause Latency to Feel Worse?

Yes.

Loss may trigger retransmissions or application recovery behavior, which can increase perceived delay.

That is why multiple metrics should be evaluated together.

What Is the Difference Between Latency and Jitter?

Latency measures delay.

Jitter measures variation in that delay.

Example:

Packets consistently arriving in 40 ms:

Low variation.

Packets arriving in:

20 ms

90 ms

35 ms

120 ms

show much greater variation.

Real time applications care about both.

How Do You Test Network Latency?

Ping can provide a basic round trip latency measurement to a reachable target.

But one test is only a snapshot.

For troubleshooting, compare:

Why Is One Ping Result Not Enough?

Latency can change throughout the day.

A test at:

4:00 PM

does not explain what happened at:

10:15 AM

Continuous monitoring creates the historical timeline needed for intermittent performance problems.

How Do You Troubleshoot High Latency?

Use a structured process:

  1. Record the exact incident time.
  2. Determine scope.
  3. Establish normal latency.
  4. Test the local gateway.
  5. Check WiFi or LAN.
  6. Check the firewall.
  7. Check the WAN.
  8. Compare multiple external destinations.
  9. Review path information.
  10. Check utilization.
  11. Review packet loss and jitter.
  12. Check failover or VPN.
  13. Review application performance.
  14. Compare historical data.
  15. Identify the probable failure domain.

Step 1: Determine the Scope

Ask:

One user?

One site?

One application?

Every internet application?

Scope immediately changes the troubleshooting path.

If one user has latency while everyone else is normal, investigate the endpoint or local access first.

Step 2: Establish What Normal Looks Like

Compare current latency with historical performance.

For example:

Normal: 18 to 25 ms

Current: 90 ms

That difference is meaningful even if the circuit remains online.

Step 3: Test the Local Gateway

If latency to the local gateway is already high, investigate:

The ISP is unlikely to be the first troubleshooting target if delay begins inside the local network.

Step 4: Compare Wired and Wireless Performance

If WiFi shows high latency but wired devices do not, investigate:

Do not blame the carrier for a local wireless problem.

Step 5: Check the Firewall

A firewall can contribute to delay if it is:

Compare firewall health with the latency incident.

Step 6: Check WAN Utilization

High circuit utilization can contribute to increased delay.

Look for correlation between:

Utilization rises

↓

Latency rises

↓

Packet loss may begin

This can indicate congestion.

Step 7: Test Multiple External Destinations

If latency rises toward one destination but not others, investigate:

If latency rises simultaneously toward many independent destinations, the likely issue may be closer to the site or common upstream carrier path.

Step 8: Review Traceroute or Path Information

Path testing can help identify:

Interpret individual hops cautiously because intermediate devices may handle diagnostic traffic differently.

Focus on end to end behavior and persistent changes that continue through the path.

What Is a Routing Change?

Internet traffic does not always follow the same path.

Carrier or upstream routing changes can alter:

A new route may introduce additional latency even though connectivity remains available.

Historical path information can help explain sudden changes.

What Is Network Congestion?

Congestion occurs when traffic demand approaches or exceeds available capacity.

Potential symptoms include:

Congestion may occur:

Fault isolation matters.

What Is Bufferbloat?

Bufferbloat occurs when excessive buffering causes large delays during congestion.

A network may maintain throughput while latency becomes very high under load.

If latency increases dramatically during uploads or downloads, queueing behavior should be investigated.

Can Uploads Cause High Latency?

Yes.

A heavily utilized upstream connection can create queueing and affect interactive traffic.

Potential sources include:

Historical utilization can help correlate these events.

Why Does Latency Get Worse at Certain Times of Day?

Recurring time patterns can indicate:

If latency rises every weekday at the same time, look for repeating demand or provider patterns.

Can a VPN Cause High Latency?

Yes.

VPN traffic may travel through:

This can increase path length and processing.

Compare:

Direct path

with:

VPN path

when appropriate.

Can a Firewall Cause High Latency?

Yes.

Investigate:

A healthy ISP cannot compensate for a struggling firewall.

Can WiFi Cause High Latency?

Yes.

Wireless latency may increase because of:

Compare wired and wireless results.

Can DNS Cause Latency?

DNS can cause slow application startup or website resolution, but that is different from sustained network path latency.

Test DNS separately from raw IP connectivity.

How Does High Latency Affect Zoom?

High network delay can affect conversational responsiveness and the overall real time experience.

But Zoom quality should be evaluated using multiple factors, including:

Use Zoom first party guidance and administrative quality information for Zoom specific interpretation.

How Does High Latency Affect VoIP?

Voice conversations are interactive.

As delay increases, users may:

Evaluate latency together with packet loss and jitter.

How Does High Latency Affect Remote Desktop?

Remote desktop requires repeated interaction between user input and remote response.

High latency can make:

feel sluggish.

How Do You Know Whether Your ISP Is Causing High Latency?

Build evidence across network boundaries.

Example:

Local gateway: normal

Firewall: normal

WAN gateway: normal or slightly elevated

Multiple external destinations: latency rises simultaneously

Historical baseline: significant deviation

That pattern can support an upstream investigation.

But avoid declaring the ISP responsible until the failure domain has been reasonably isolated.

What Evidence Should You Give Your ISP?

Provide:

Specific evidence produces a better escalation.

Example High Latency ISP Escalation

> Location 22 normally averages approximately 24 ms toward our monitored > external targets. Beginning at 2:11 PM, latency increased above 110 ms > across multiple independent destinations while the local gateway and > firewall remained stable. Similar events occurred yesterday at 1:55 PM > and Monday at 2:07 PM. Please investigate circuit and upstream path > performance during these timestamps.

This is more useful than:

"The internet is slow."

When Is High Latency Not the ISP's Fault?

Examples include:

Good troubleshooting isolates the problem before assigning responsibility.

Can SD WAN Reduce High Latency?

SD WAN can select alternate eligible paths when configured policies identify better network conditions.

However, it does not repair the underlying carrier path.

If traffic repeatedly moves away from one circuit because of latency, that circuit should still be investigated.

Should High Latency Trigger WAN Failover?

It can be considered in performance based routing designs depending on platform capabilities and application requirements.

Thresholds should be carefully designed and tested.

Overly aggressive policies can create unnecessary path changes.

What Is Latency Monitoring?

Latency monitoring measures network delay continuously over time.

A useful system should help answer:

Why Is Historical Latency Monitoring Important?

Because current performance may be normal by the time a technician investigates.

Historical data can show:

Before incident: 22 ms

During incident: 140 ms

After incident: 24 ms

Without history, that evidence disappears.

Why Should You Monitor More Than Average Latency?

Averages can hide short but severe events.

Consider:

A five minute latency spike during a critical customer call may matter even if the daily average remains acceptable.

How Can ADAM Pulse Help Troubleshoot High Latency?

ADAM Pulse should provide the historical network context needed to determine:

What is normal for this location?

When did latency increase?

Was the LAN healthy?

Was the firewall healthy?

Which WAN was active?

Were multiple destinations affected?

Did packet loss or jitter also change?

Has this happened before?

That allows USA Telecom to troubleshoot from evidence rather than user perception alone.

The ADAM Pulse Latency Troubleshooting Framework

BASELINE → TIME → SCOPE → LOCAL → EDGE → WAN → PATH → APPLICATION → HISTORY

Baseline

Know normal latency.

Time

Identify the incident period.

Scope

Determine who and what is affected.

Local

Validate LAN and WiFi.

Edge

Validate firewall health.

WAN

Evaluate the carrier connection.

Path

Compare destinations and routes.

Application

Determine whether the service itself is affected.

History

Identify recurrence and trends.

Do Not Ask Only "Is the Internet Slow?"

Ask:

> Where did the delay begin?

That question turns a vague performance complaint into a structured network investigation.

ADAM Pulse provides managed network monitoring designed to help USA Telecom customers establish latency baselines, detect meaningful performance changes, preserve historical evidence and improve network and carrier troubleshooting.

Know normal.

Detect change.

Follow the path.

Preserve the evidence.

Escalate intelligently.

Talk with USA Telecom about using ADAM Pulse to monitor latency across business locations, WAN connections and critical applications.

Frequently asked questions

What causes high network latency?

Common causes include geographic distance, congestion, poor routing, WiFi, firewall processing, VPN architecture, ISP problems and saturated network connections.

Is latency the same as internet speed?

No. Internet speed generally describes bandwidth or throughput, while latency measures delay.

How do I troubleshoot high latency?

Compare current performance against a baseline, test the local gateway, LAN, firewall and WAN, compare multiple external destinations, review utilization and path information, and examine historical monitoring.

How do I know whether my ISP is causing high latency?

First verify that local gateway, LAN and firewall performance are normal. Then compare multiple external destinations and historical data to determine whether increased delay begins beyond the local network.

Can high latency cause Zoom and VoIP problems?

Yes. High delay can affect conversational responsiveness and interactive application performance. Packet loss and jitter should also be evaluated.

Sources

Editorial note

A single test from a single location at a single moment rarely proves where a fault sits. Correlate against history, test from more than one point, and preserve timestamped evidence before changing configuration or rebooting equipment.

See how ADAM Pulse localises latency to LAN, firewall or ISP

ADAM Pulse measures latency continuously across the path, so the evidence exists before anyone reboots the firewall or opens a carrier ticket.

Related articles

Latency vs jitter vs packet loss · Why is my network slow when the speed test is fast? · Is it the ISP or the firewall? · How to prove your ISP is the cause · Network Monitoring topic hub

← More from the ADAM Pulse Knowledge Base

Prepared by ADAM Pulse (USA Telecom Consulting LLC) — managed AI governance, network, and NOC services. Support: (888) 989-4872 · support@adampulse.us