A Wi-Fi icon, a successful ping, or a visible route does not prove that the application you need can reach its TCP service. The reliable way to troubleshoot a slow or unreliable TCP/IP connection is to test one layer at a time: define the symptom, verify the local adapter and gateway, separate DNS from routing, test the actual destination port, and capture traffic when the failure can be reproduced. Compare one variable at a time—another device, Ethernet instead of Wi-Fi, a hostname instead of its IP address, or a local gateway instead of a remote server—while recording times and exact results.
The procedures below focus on Windows 10, Windows 11, and Windows Server. Labels and command output can differ by edition, policy, adapter driver, and network design. ICMP may be filtered, and packet captures can contain sensitive addresses and data, so follow your organization’s rules before collecting or sharing them.
As an Amazon Associate I earn from qualifying purchases.
Start by defining what is actually failing
“Slow internet” can describe several different faults. Write down the exact symptom before changing anything:
- Slow transfer: a download, upload, file copy, or page takes longer than the same workload normally takes.
- High delay: interactive applications feel laggy even though they eventually respond.
- Intermittent drops: an established session disconnects, Wi-Fi disappears, or traffic works again after a short pause.
- Connection establishment failure: a client cannot open a session to a host and port.
- Application timeout: a connection appears present, but a particular program cannot complete an operation.
- Name-resolution failure: a hostname fails while a known IP address may still work.
Record the application, destination hostname and IP address if known, destination port, wired or Wi-Fi connection, local time and time zone, and whether other devices or services are affected. Record the exact command and output rather than writing only “ping failed.” Use the same destination and workload when comparing results; a speed test to one server is not a substitute for testing the service that is slow.
#1 Best Overall
- DUAL-BAND WIFI 6 ROUTER: Wi-Fi 6(802.11ax) technology achieves faster speeds, greater capacity and reduced network congestion compared to the previous gen. All WiFi routers require a separate modem. Dual-Band WiFi routers do not support the 6 GHz band.
- AX1800: Enjoy smoother and more stable streaming, gaming, downloading with 1.8 Gbps total bandwidth (up to 1200 Mbps on 5 GHz and up to 574 Mbps on 2.4 GHz). Performance varies by conditions, distance to devices, and obstacles such as walls.
- CONNECT MORE DEVICES: Wi-Fi 6 technology communicates more data to more devices simultaneously using revolutionary OFDMA technology
- EXTENSIVE COVERAGE: Achieve the strong, reliable WiFi coverage with Archer AX1800 as it focuses signal strength to your devices far away using Beamforming technology, 4 high-gain antennas and an advanced front-end module (FEM) chipset
- OUR CYBERSECURITY COMMITMENT: TP-Link is a signatory of the U.S. Cybersecurity and Infrastructure Security Agency’s (CISA) Secure-by-Design pledge. This device is designed, built, and maintained, with advanced security as a core requirement.
Draw the path when the network is managed
For a business, school, or hosted environment, sketch the client-to-server path. Include the client adapter, wireless access point, switch, default gateway, firewall, VPN or inspection device, WAN links, remote firewall, and server. Mark where you can collect observations. Microsoft’s Windows Server troubleshooting guidance recommends a network diagram and traces because the apparent problem may be several hops away from the computer reporting it.
Check the Windows device and local configuration
Inspect status before resetting anything
- Open Settings > Network & Internet and select Wi-Fi or Ethernet. Confirm that the intended adapter is connected and that Windows is not using an unexpected VPN, proxy, or metered connection.
- Run the built-in network troubleshooter when it is available and record its findings. Treat its result as a clue, not a complete diagnosis.
- Open Command Prompt and run
ipconfig. Note the active adapter, IPv4 or IPv6 address, subnet mask or prefix, default gateway, and DNS servers.
An IPv4 address beginning with 169.254 usually means Windows assigned an automatic private address because it did not obtain a DHCP lease. First determine whether the network is supposed to use automatic addressing. A deliberately configured static network can make a 169.254 observation misleading, and changing a static address on a managed network can create a new outage. If DHCP is expected, check the access point, cable, switch port, and DHCP service before trying repeated resets.
Check adapter, cable, and proxy details
Reconnect the adapter and verify that the Ethernet cable is seated in the correct router or switch port. For a wired problem, try a known-good cable or another device on the same port. Review proxy settings and DNS configuration if the application uses a hostname or if browser traffic behaves differently from other programs. Record the existing values before changing them; replacing DNS servers or static settings without understanding the intended design can hide the original fault and violate a managed-network policy.
Test the default gateway, with the ICMP limitation in mind
Use the gateway address shown by ipconfig:
ping <default-gateway>
A reply proves that the computer exchanged ICMP packets with that gateway at that moment. It does not prove that DNS works, that the Internet route is usable, that a remote host is reachable, or that an application is listening on its TCP port. No reply is also not conclusive: a gateway, firewall, or host may filter or deprioritize ICMP. Microsoft’s official guidance puts it plainly: “Ping is useful to test basic connectivity. However, you shouldn’t rely upon it to prove overall connectivity.”
Separate a Wi-Fi or LAN fault from an Internet-path fault
Compare another device and another connection type
Test another device on the same Wi-Fi network against the same destination. If several devices show the same failure, suspect the access point, router, ISP link, or an upstream service rather than one Windows installation. If only one computer is affected, concentrate on its adapter, driver, address, VPN, proxy, or local security software.
Rank #2
- Dual-band Wi-Fi with 5 GHz speeds up to 867 Mbps and 2.4 GHz speeds up to 300 Mbps, delivering 1200 Mbps of total bandwidth¹. Dual-band routers do not support 6 GHz. Performance varies by conditions, distance to devices, and obstacles such as walls.
- Covers up to 1,000 sq. ft. with four external antennas for stable wireless connections and optimal coverage.
- Supports IGMP Proxy/Snooping, Bridge and Tag VLAN to optimize IPTV streaming
- Access Point Mode - Supports AP Mode to transform your wired connection into wireless network, an ideal wireless router for home
- Advanced Security with WPA3 - The latest Wi-Fi security protocol, WPA3, brings new capabilities to improve cybersecurity in personal networks
When Wi-Fi is suspect, connect the computer directly to the modem with Ethernet if that arrangement is practical and permitted. Microsoft’s Windows Wi-Fi procedure uses this comparison: if the direct wired path reaches the Internet while the computer cannot reliably reach the Wi-Fi router, the Wi-Fi router or wireless portion is implicated. This is an isolation test, not proof that new hardware is required. For a wired fault, compare another cable, another router port, and another device.
Use a simple comparison matrix
| Comparison | If the second test works | What it suggests |
|---|---|---|
| One computer versus another on the same network | Only one computer fails | Local adapter, address, driver, VPN, proxy, or endpoint software |
| Wi-Fi versus direct Ethernet | Ethernet works | Wireless router, radio conditions, or Wi-Fi configuration |
| Local gateway versus remote destination | Gateway works but remote host fails | DNS, route, firewall, ISP path, remote policy, or service |
| Hostname versus known IP address | IP works but hostname fails | Name resolution or name-based policy |
| Ping versus application port | Ping works but port fails | Service, firewall, or TCP policy rather than basic ICMP reachability |
| Source-side versus destination-side capture | Packets stop between observation points | Path or intermediary device; if they arrive but no reply leaves, inspect the destination |
Determine whether DNS, routing, or the service is responsible
Check name resolution without changing it prematurely
If a hostname fails but a known IP address works, investigate DNS. Confirm that a DNS server is listed in ipconfig and compare the result with the network’s intended configuration. If both the name and the IP fail, continue with route and service tests. Do not switch providers or hard-code a DNS address merely because one lookup was slow; first capture the current configuration and establish whether the problem is repeatable.
Read tracert as a route clue, not a verdict
Run:
tracert <destination>
Windows sends probes with increasing time-to-live values and displays responses from intermediate routers. Some routers do not return the diagnostic messages that tracert expects, so an asterisk can appear even while that router forwards application traffic normally. A single missing hop, especially in the middle of an otherwise complete route, is not proof that the hop is dropping your connection. Look for a pattern that persists across repeated tests and compare it with an endpoint-level test.
Use pathping for repeated loss and latency clues
Run:
pathping <destination>
pathping sends multiple probes over time and calculates latency and loss observations for intermediate hops. The output is useful for finding a consistent change after a particular point, but it remains an ICMP-based clue. A router may prioritize or suppress diagnostic replies differently from application packets. Corroborate a suspicious hop with the actual service test and, for persistent incidents, a packet capture.
Test the destination TCP port
A service can be reachable by ICMP while its TCP port is closed, filtered, or unhealthy. Test the port you are authorized to test with a port-specific tool. For example, Telnet can attempt a TCP connection:
Rank #3
- NIGHTHAWK WIFI 6 ROUTER FOR YOUR WHOLE HOME: Delivers fast, reliable WiFi across every room of your apartment or small home for streaming, gaming, video calls, and smart home devices, all running at the same time without slowing each other down.
- WORKS WITH YOUR EXISTING INTERNET SERVICE: Pairs with your existing modem or gateway via ethernet. Compatible with most cable, fiber, DSL, and satellite providers. Some gateways and modem router combos may require bridge mode. No coax needed.
- SET UP AND MANAGE YOUR NETWORK WITH THE NIGHTHAWK APP: Download the free Nighthawk app on iOS or Android for guided setup. Manage WiFi, run speed tests, pause devices, and set up guest networks from anywhere. Active internet required.
- READY FOR THE DEVICES YOU ALREADY OWN: Your phones, laptops, and TVs work right out of the box. WiFi 6 delivers speeds up to 1.8 Gbps across 2.4 GHz and 5 GHz bands. Backward compatible with WiFi 5 and earlier.
- COVERAGE IN EVERY ROOM: Covers up to 1,500 sq. ft. for up to 20 connected devices. Walls, floors, and interference can reduce range. Larger or multi-story homes may benefit from a NETGEAR Orbi mesh WiFi system.
telnet <server> <port>
Microsoft also documents PsPing for port checks, for example:
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →psping -n 4 <server>:<port>
A successful TCP connection shows that a listener responded at that moment; it does not prove that the application transaction, authentication, or data transfer will succeed. If the port is not listening, investigate the service. If packets are silently dropped, inspect host firewalls, network firewalls, security groups, VPN policy, and any inspection device along the path.
Investigate TCP stalls, retransmissions, and resets with packet evidence
What common packet patterns mean
TCP is connection-oriented and retransmits when expected acknowledgments or responses do not arrive. Repeated SYN packets can indicate difficulty establishing a connection. Repeated data segments or missing acknowledgments during a transfer can indicate loss, a delayed response, or a receiver that cannot process traffic promptly. A TCP reset can be generated by an endpoint, an application, or an intermediary policy. None of these patterns, by itself, identifies a faulty router.
Packet-size or MTU complications are another possibility, particularly when a VPN, tunnel, or path handles larger packets differently. Treat MTU as a hypothesis supported by evidence, not as a reason to change settings immediately. A reset or retransmission may also reflect a remote policy decision or a service that closed the connection intentionally.
Capture while reproducing the failure
- Choose one affected client, one destination, and one repeatable operation. Note the precise start and end times, including time zone.
- Capture traffic on the source while starting the operation. Preserve the destination address, TCP port, and any error shown by the application.
- If you control the destination, capture there at the same time. Align timestamps, source and destination addresses, ports, and TCP sequence information.
- Compare what appears at each endpoint. If a packet leaves the source but never appears at the destination, the loss is between those observation points. If it arrives at the destination but no response leaves, examine the destination host, firewall, listener, or application.
- Stop the capture after the failure and store the original file securely. Share it only under your organization’s privacy and security rules.
Two-sided traces are more informative than a single capture because they show where a packet disappears. They still require context: offloading, virtualization, load balancers, asymmetric routing, and capture-point differences can make timestamps and packet counts look inconsistent.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallCrashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteRank #4
- 𝐅𝐮𝐭𝐮𝐫𝐞-𝐑𝐞𝐚𝐝𝐲 𝐖𝐢-𝐅𝐢 𝟕 - Designed with the latest Wi-Fi 7 technology, featuring Multi-Link Operation (MLO), Multi-RUs, and 4K-QAM. Achieve optimized performance on latest WiFi 7 laptops and devices, like the iPhone 16 Pro, and Samsung Galaxy S24 Ultra.
- 𝟔-𝐒𝐭𝐫𝐞𝐚𝐦, 𝐃𝐮𝐚𝐥-𝐁𝐚𝐧𝐝 𝐖𝐢-𝐅𝐢 𝐰𝐢𝐭𝐡 𝟔.𝟓 𝐆𝐛𝐩𝐬 𝐓𝐨𝐭𝐚𝐥 𝐁𝐚𝐧𝐝𝐰𝐢𝐝𝐭𝐡 - Achieve full speeds of up to 5764 Mbps on the 5GHz band and 688 Mbps on the 2.4 GHz band with 6 streams. Enjoy seamless 4K/8K streaming, AR/VR gaming, and incredibly fast downloads/uploads.
- 𝐖𝐢𝐝𝐞 𝐂𝐨𝐯𝐞𝐫𝐚𝐠𝐞 𝐰𝐢𝐭𝐡 𝐒𝐭𝐫𝐨𝐧𝐠 𝐂𝐨𝐧𝐧𝐞𝐜𝐭𝐢𝐨𝐧 - Get up to 2,400 sq. ft. max coverage for up to 90 devices at a time. 6x high performance antennas and Beamforming technology, ensures reliable connections for remote workers, gamers, students, and more.
- 𝐔𝐥𝐭𝐫𝐚-𝐅𝐚𝐬𝐭 𝟐.𝟓 𝐆𝐛𝐩𝐬 𝐖𝐢𝐫𝐞𝐝 𝐏𝐞𝐫𝐟𝐨𝐫𝐦𝐚𝐧𝐜𝐞 - 1x 2.5 Gbps WAN/LAN port, 1x 2.5 Gbps LAN port and 3x 1 Gbps LAN ports offer high-speed data transmissions.³ Integrate with a multi-gig modem for gigplus internet.
- 𝐎𝐮𝐫 𝐂𝐲𝐛𝐞𝐫𝐬𝐞𝐜𝐮𝐫𝐢𝐭𝐲 𝐂𝐨𝐦𝐦𝐢𝐭𝐦𝐞𝐧𝐭 - TP-Link is a signatory of the U.S. Cybersecurity and Infrastructure Security Agency’s (CISA) Secure-by-Design pledge. This device is designed, built, and maintained, with advanced security as a core requirement.
Measure slowness instead of guessing
Keep throughput, latency, and loss separate
Record transfer rate, response time, and packet loss as different observations. A high throughput result does not guarantee low interactive delay, and a low throughput result does not identify the limiting component. Measure the same endpoints, workload, path, and relevant system conditions for a baseline and for the affected period. Include whether the test used Wi-Fi or Ethernet, a VPN, a busy server, or a different time of day.
Compare like with like
- Repeat the same file transfer or application operation, not just a generic speed test.
- Run a local-network comparison when possible, such as a server on the same LAN, to separate LAN limits from Internet-path limits.
- Compare an idle period with the failure period and record concurrent uploads, backups, video calls, or large downloads.
- Check whether only one protocol or application is slow. A single service can be overloaded or rate-limited while other TCP services remain normal.
Windows Server performance guidance discusses high-latency and high-bandwidth scenarios, retransmissions, and features such as Receive Side Scaling and Virtual Machine Queue. Adapter or operating-system tuning should be considered only when measurements and the platform configuration support it. Changing RSS, VMQ, MTU, auto-tuning, or offload settings without a baseline can reduce performance or create a second problem.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.A decision path for common outcomes
The computer cannot reach its gateway
Check the adapter state, address, subnet, cable, access-point association, and switch or router port. If DHCP is expected, investigate lease acquisition and the device providing DHCP. Compare another device and, for Wi-Fi, perform the direct Ethernet test. If ICMP is filtered, use a permitted local service or management check rather than concluding that the link is down from one failed ping.
The gateway works, but the hostname does not resolve
Inspect DNS server entries and proxy or VPN settings. Compare a hostname lookup with a known IP address and note whether the failure affects every name or only one domain. Preserve the original configuration before making a DNS change.
Free tools Windows power users keep installed
One-click scans. No signup required.
The name resolves, but the TCP port fails
Test the exact port with an authorized port tool. Check whether the destination service is listening, then inspect host and network firewall policy, VPN rules, security groups, and remote access restrictions. A successful ping does not override a closed or filtered TCP port.
Best Value
- Dual band router upgrades to 1200 Mbps high speed internet (300mbps for 2.4GHz plus 900Mbps for 5GHz), reducing buffering and ideal for 4K stream
- Full Gigabit Ports - Gigabit Router with 4 Gigabit LAN ports, ideal for any internet plan and allow you to directly connect your wired devices
- Boosted Coverage - Four external antennas equipped with Beamforming technology extend and concentrate the Wi-Fi signals
- MU-MIMO technology - (5GHz band) allows high speeds for multiple devices simultaneously
- Access Point Mode - Supports AP Mode to transform your wired connection into wireless network, an ideal wireless router for home
The port opens, but the application times out or transfers slowly
Capture the transaction while reproducing it. Look for retransmissions, delayed acknowledgments, resets, and packets that appear at one endpoint but not the other. Check the application and server logs at the recorded time, because a TCP session can be healthy while the service is overloaded, waiting on storage, or rejecting a request.
Several devices fail at the same time
Compare the local gateway, another external destination, and the affected service. Inspect the router, firewall, WAN link, VPN, and provider status through the channels available to you. A route change or upstream outage is more plausible than a simultaneous adapter fault, but packet evidence and service tests are still needed to locate the boundary.
Common mistakes that make diagnosis harder
- Treating the Wi-Fi icon as proof of Internet access: association with an access point does not test DNS or the destination service.
- Declaring a broken router from an asterisk in tracert: intermediate routers may suppress diagnostic replies while forwarding traffic.
- Using one ping as the entire investigation: ICMP can be filtered, and ping does not test a TCP listener.
- Changing several settings at once: a new DNS server, driver, MTU, firewall rule, and cable provide no reliable way to know which change mattered.
- Assuming every retransmission identifies the culprit: retransmission proves that expected TCP progress was not observed, not where the cause lies.
- Sharing an unredacted capture: traces may contain internal addresses, hostnames, usernames, URLs, and application data.
- Buying hardware as a first response: a known-good Ethernet cable has a narrow role in a wired-versus-Wi-Fi isolation test or in replacing a suspect cable. It cannot repair an ISP route, DNS policy, TCP service, or overloaded application. If a computer has no Ethernet port, a compatible USB-to-Ethernet adapter can enable the comparison.
Prepare an escalation package
When the issue persists, give the next administrator or provider a reproducible record:
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- Start and end times with time zone, affected users or devices, and whether the failure is continuous or intermittent.
- Application, destination hostname and IP address, destination TCP port, and the exact user action that fails.
- Wired or Wi-Fi status, VPN or proxy state, adapter address, gateway, and DNS configuration from
ipconfig. - Results from the gateway test, hostname-versus-IP comparison,
tracert,pathping, and the authorized port check. - A network diagram showing the observation points and any firewall, inspection, tunnel, or WAN device in the path.
- Packet captures taken during a repeatable failure, ideally at both endpoints, with sensitive information handled according to policy.
- What changed, what was tested, and which single variable was changed between comparisons.
This record lets an administrator correlate packet behavior with firewall, router, server, and application logs instead of repeating broad resets. It also preserves the distinction between a local-link problem, an Internet-path problem, a TCP exchange problem, and an application that is slow after TCP has already connected.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
