A VPS can feel slow even when its advertised bandwidth is high. “Slow” may mean high latency, packet loss, DNS delay, application processing, a one-way throughput limit, or contention among CPU, disk and virtual-network resources. Diagnose those symptoms separately: define the failing direction and destination, verify the VM locally, map the network path, measure controlled throughput, then check resource and plan limits before changing firewall, route or MTU settings.
Need a server to try this on? Antihost offers Linux VPS plans with NVMe storage, full root access and a 14-day free trial.
What “slow VPS network” actually means
Start by replacing the single word slow with a measurable symptom. A website may take a long time to begin responding because of DNS or application work while its bulk transfer rate is normal. SSH or RDP can feel delayed because of round-trip latency or packet loss even when CPU usage is low. A file transfer can have excellent latency but poor throughput because of a VM egress ceiling, a shared virtual port, disk I/O or a congested path.
As an Amazon Associate I earn from qualifying purchases.
Recommended Free Tools
- Latency: the round-trip time before a packet is answered. It affects interactive SSH, RDP, APIs and game traffic.
- Jitter: variation in latency. Bursty delay makes voice, video and remote desktops feel erratic.
- Packet loss: missing packets trigger retransmissions and can collapse TCP throughput.
- DNS delay: a hostname lookup is slow or fails while direct access to the IP works.
- Throughput: the sustained upload or download rate once a transfer is underway.
- Application delay: the service itself is slow even though network tests are healthy.
Record the direction (upload, download or both), destination, protocol, application, time, VPS region, client network and a comparison endpoint that normally performs well. A single browser speed test cannot tell whether the limiting factor is your client, the route, the VM or the service.
Run these checks in order
1. Verify the VM’s address, route and listening service
Use the provider’s out-of-band console before changing network settings. Console access is your recovery path if a firewall or route change cuts off SSH.
#1 Best Overall
- Read Before You Buy — No Video Output: These adapters support charging and USB 2.0 data transfer, but cannot transmit video signals. Except for standard USB webcams (which use USB data only), they are not compatible with HDMI/DisplayPort cables, video-capable USB-C hubs, or docking stations with video output.
- Convert USB-A Ports to USB-C: Designed to connect USB-C earphones, cables, flash drives, card readers, and other USB-C accessories to standard USB-A ports. Plug-and-play with no drivers or software required.
- Aluminum Alloy Housing: Built with a sturdy aluminum alloy shell that aids in heat dissipation and protects against daily wear and scratches. Designed to maintain a stable and secure connection.
- Compact & Travel-Friendly: The ultra-compact design allows the adapter to stay plugged into your device without blocking adjacent ports or adding bulk, reducing wear and tear on your original USB ports.
- 12-Month Warranty: Backed by a 12-month manufacturer warranty for peace of mind. Designed to meet strict quality control standards for reliable everyday performance.
- Run
ip addrand confirm that the expected IPv4 and, if applicable, IPv6 addresses are attached to the interface. - Run
ip routeand confirm a default route points to the provider’s gateway. - Inspect the active firewall with the tool your distribution uses:
ufw status,iptables -L -n -vornft list ruleset. Look for drops on the service port and for rules that rate-limit or reject return traffic. - Confirm that the application is actually listening:
ss -tlnp. Check that it is bound to the intended address and port rather than only to127.0.0.1.
A missing default route, an incorrect address, or an over-broad firewall rule can look like a bandwidth problem. Keep the console open while testing any rule or route change, and make one reversible change at a time.
2. Separate DNS from IP reachability
Test an IP address and a hostname independently:
ping 1.1.1.1
ping google.comIf the IP responds but the hostname does not, inspect the resolver configuration and the DNS server’s reachability. For an actual service, use its protocol rather than relying only on ICMP; for example, request the HTTPS endpoint with curl and note whether the delay occurs before connection, during TLS setup or while the response body is generated. An IP test that succeeds does not prove that the application, DNS, or TCP port is healthy.
3. Map the path to the affected destination
Run ping first, then use traceroute or mtr to the same destination. For an HTTPS service, test the protocol and port being evaluated:
mtr DESTINATION -T -P 443 -c 100 --reportMTR and traceroute show each hop and its round-trip time. TCP probes are more representative for a TCP service than ICMP probes when intermediate routers deprioritize or block ICMP. A hop that shows loss or does not answer is not, by itself, proof of end-to-end loss: some routers simply suppress diagnostic replies. Treat a hop as a problem only when the loss or increased latency continues through subsequent hops and reaches the destination.
Compare paths for more than one destination and, when relevant, more than one protocol. A route can be healthy to one network and congested or poorly peered to another, which explains why one website is slow while another is normal.
4. Measure controlled throughput with iperf3
Use two controlled endpoints rather than a public web speed-test server. Install compatible iperf3 versions, permit the test port in both firewalls, and place one endpoint outside the VPS’s provider network when you are testing Internet egress.
Rank #2
- 5-in-1 USB-C Hub: Experience comprehensive connectivity featuring a Power Delivery input, two USB-A 2.0 ports, a USB-A 3.0 port, and an HDMI port. (Note: The USB-C power delivery input port is only for connecting an external wall charger to power your laptop and cannot power peripheral devices.)
- 90W Pass-Through Charging: Achieve optimal charging with 90W pass-through power to your laptop, supported by a total input of 100W, with the hub reserving 10W for operational efficiency. (Note: Wall charger not included.)
- Quick Data Transfers: Accelerate your productivity with rapid data transfers using a high-speed 5Gbps USB 3.0 port and two 480Mbps USB 2.0 ports.
- 4K HDMI Display: Enhance your visual experience with a hub capable of delivering 4K resolution at 30Hz in both mirror and extend modes. Please note that this hub is compatible with MacBook (macOS 12 and newer), Windows 10 and 11, ChromeOS, and laptops equipped with DP Alt Mode and Power Delivery. Note: This device is not compatible with Linux.
- What You Get: Anker USB-C Hub (5-in-1, 4K HDMI), welcome guide, 18-month warranty, and our friendly customer service.
# On the test server
iperf3 -s
# On the VPS (client)
iperf3 -c SERVERRepeat the test in both directions and with multiple streams:
iperf3 -c SERVER -R
iperf3 -c SERVER -P 4
iperf3 -c SERVER --bidir-Rreverses the test, so the server sends to the VPS.-P 4runs four parallel streams; compare it with a single stream to expose per-flow or TCP-window limits.--bidirmeasures simultaneous traffic in both directions and can reveal contention that separate tests miss.
Run several intervals at different times and save the output with the destination, port, region and command. A low single-stream result but a much higher multi-stream result often indicates per-flow behavior or latency rather than a hard port ceiling. Conversely, low results in every mode suggest a VM, provider or path limit. Throughput limitations often come from VM or disk limits rather than the network itself.
Interpret direction carefully. A normal forward test only proves that traffic can travel from the client to the server under those conditions; it says nothing about the reverse path. A reverse test can expose asymmetric routing, a directional egress policy or a congested peering link. Simultaneous bidirectional traffic is useful when the real workload uploads and downloads at once, but it should not replace isolated tests because each direction can compete for the same virtual-port or host capacity.
Check VM and host resources while the test runs
Watch the VPS during mtr and iperf3, not only before or after them. Correlate a throughput drop or latency spike with the following counters:
Crashes, 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 minutePC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11- CPU and steal time: exhausted vCPUs, high load or high steal time can prevent the network stack and encryption from keeping up.
- Memory and swap: swapping adds delay and can make an otherwise healthy connection appear intermittent.
- Disk latency and I/O wait: a web server, database or file transfer may be disk-bound. Persistent-disk activity can interfere with VM network throughput.
- Virtual NIC counters: inspect received/transmitted bytes, errors, drops and saturation in the guest and in the provider’s graphs.
- Connection and queue pressure: a full listen queue, excessive concurrent connections or application rate limiting can limit useful throughput without showing a physical network fault.
Cloud guidance identifies vCPU exhaustion and virtual-NIC saturation as causes of latency and loss. If resource graphs rise at the same time as the network symptom, reproduce the test after reducing application load or moving the workload to an idle VM. If the symptom disappears, scaling CPU, disk performance or the VM’s network tier may solve it more reliably than changing routes.
Take observations during both a good period and a bad period. Averages taken after a transfer finishes can hide short bursts of steal time, queue growth or packet drops. Compare provider-side graphs with guest counters: a quiet guest interface alongside a busy provider port suggests an accounting or host-level limit, while drops visible in both places point toward saturation or congestion. Keep the test workload simple so that database queries, compression and encryption do not obscure the network result.
Rank #3
- Sleek 7-in-1 USB-C Hub: Features an HDMI port, two USB-A 3.0 ports, and a USB-C data port, each providing 5Gbps transfer speeds. It also includes a USB-C PD input port for charging up to 100W and dual SD and TF card slots, all in a compact design.
- Flawless 4K@60Hz Video with HDMI: Delivers exceptional clarity and smoothness with its 4K@60Hz HDMI port, making it ideal for high-definition presentations and entertainment. (Note: Only the HDMI port supports video projection; the USB-C port is for data transfer only.)
- Double Up on Efficiency: The two USB-A 3.0 ports and a USB-C port support a fast 5Gbps data rate, significantly boosting your transfer speeds and improving productivity.
- Fast and Reliable 85W Charging: Offers high-capacity, speedy charging for laptops up to 85W, so you spend less time tethered to an outlet and more time being productive.
- What You Get: Anker USB-C Hub (7-in-1), welcome guide, 18-month warranty, and our friendly customer service.
Check the exact plan and provider limits
Do not compare your result with a headline “gigabit” figure without checking the exact VM size and region. Confirm:
- the instance’s documented egress and ingress ceilings;
- whether the limit is per VM, per virtual port, per flow or shared among tenants;
- regional and peering differences for the destination you tested;
- traffic-shaping, fair-use and egress policies;
- whether private networking has a separate capacity or guarantee.
Many virtual servers use a shared network port. Guaranteed throughput or private networking may require confirmation before ordering, and any published figure is specific to the plan and location. Ask support to state the expected sustained rate, burst behavior, directional limits and whether the value applies to one stream or aggregate traffic.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
When evaluating a different VPS, compare latency and jitter to your real users, packet loss by destination and protocol, single-stream and multi-stream throughput, upload versus download, vCPU and egress limits, disk interference, MTU behavior, region and peering, shared-port or fair-use terms, and the provider’s ability to interpret MTR reports.
How to test whether a provider limit is involved
“Is the provider throttling me?” is answerable only with a controlled comparison. First test several destinations, because a single congested route can imitate throttling. Then repeat at different times, in both directions, with one stream and multiple streams. A ceiling that appears at nearly the same rate across unrelated destinations and persists when the VM is idle is consistent with a virtual-port, egress or fair-use limit. A problem limited to one destination, one protocol or one time window points more toward peering, path congestion or remote-service limits.
Compare the observed behavior with the exact plan terms, not a marketing headline for another instance class. Ask support whether the advertised rate is burst capacity, an aggregate cap, a per-flow value or a best-effort estimate. Request confirmation for the tested region and whether traffic between the provider’s own facilities is treated differently from Internet egress. Keep the commands, timestamps and raw outputs so support can correlate them with host and port statistics.
Investigate MTU and fragmentation only when symptoms fit
MTU problems usually produce a pattern rather than a universal slowdown: inconsistent UDP, large transfers that stall, or failures to particular networks while small packets work. A path with a smaller MTU can fragment packets, delay them or reorder fragments. Encapsulation used by tunnels and private networks can make the effective path MTU smaller than the interface default.
Rank #4
- Dual Converters, Infinite Potential：Includes 2× USB C male to USB A female adapters and 2× USB A male to USB C female adapters. Perfect for a wide range of uses—tablets with Bluetooth keyboards, expand USB ports on macbook, and more. Two different converters for all your daily needs
- Next-Level 10Gbps & 3A Charging: No more slow 480Mbps, this usb to usb c adapter has a transfer speed of up to 10Gbps, allowing you to do more transferring in less time. This usb adapter fits both USB A and USB C charger, supporting up to 3A fast charging
- Upgraded Exquisite Craftsmanship: With an aluminum alloy housing and metal connector, the usbc to usb adapter is extremely durable and sturdy. Rigorously tested to withstand more than 10,000 times of plugging and unplugging, ensuring long-lasting performance
- Broad Compatible: The usb c to usb adapter widely supports all USB C/ USB A devices like laptops, tablets, cellphones, car chargers, and phone chargers. Such as compatible with MacBook Pro/Air 2023/2022, Thunderbolt 4/3 Devices,Apple MagSafe Watch 9/8/7/SE/Ultra, iPad Pro 2022/2021, Samsung Galaxy S23/S20/S10, and iPhone 17/16/15 Pro. Plug and play
- Please Note: To reach 10Gbps speed, keep the cable under 3.3 ft. For USB A Male to USB C adapters, try flipping the USB C connector. USB C Male to USB A adapters support bidirectional 10Gbps transfer within 3.3 ft
- Compare small and large requests to the same destination and test both UDP and TCP when the application uses both.
- Inspect the interface MTU and the path behavior before changing it.
- Make any MTU change from the provider console, record the original value, and prepare a rollback command.
Do not lower MTU as a general speed tweak. An incorrect value can reduce efficiency or break connectivity to other networks. Change it only when packet-size tests and path evidence support the diagnosis.
Common symptom patterns and their likely causes
| Observed pattern | What to test next |
|---|---|
| SSH or RDP is laggy while CPU looks normal | Run ping and TCP MTR to the client or service; check jitter, sustained loss and route changes. |
| One website is slow but other destinations are fast | Compare MTR and traceroute paths, DNS resolution and TCP tests to each destination. |
| Downloads are fast but uploads are slow (or the reverse) | Run iperf3 normally and with -R; check directional egress policy and virtual-port limits. |
| One stream is slow, several streams are fast | Compare single-stream and -P tests; investigate per-flow limits, TCP windowing and latency. |
| Throughput falls when backups or databases run | Correlate disk latency, I/O wait, CPU and NIC counters during the transfer. |
| Large or UDP transfers fail while small packets work | Check MTU, fragmentation and protocol-specific firewall rules. |
| IP access works but hostname access does not | Inspect resolver configuration and DNS reachability rather than changing bandwidth settings. |
Fix the cause instead of applying generic “speed tweaks”
Match the remedy to the measured failure. Resolver delay calls for correcting DNS configuration or reachability, not raising the VM’s bandwidth. Persistent path loss or high latency may require a different region, network path or provider rather than a larger disk. A VM that reaches its vCPU, disk or virtual-NIC ceiling needs a size or workload change. A confirmed shared-port or fair-use cap requires a plan with a different entitlement or a provider that offers guaranteed bandwidth; kernel tuning cannot remove a provider-side ceiling.
Change one variable at a time and rerun the same tests. Moving regions, changing MTU, altering firewall rules and resizing the VM simultaneously makes it impossible to identify what helped. Preserve the original outputs, configuration and timestamps so you can revert a change or show support the before-and-after behavior.
What not to conclude from a single test
- A browser speed test is not a guaranteed VPS capacity measurement; it includes the client, browser, selected server, route and application protocol.
- An unresponsive traceroute hop is not proof that packets are being dropped for the final service.
- Low CPU in the guest does not rule out a vNIC ceiling, host contention, disk interference or a provider traffic policy.
- A high result to a nearby test server does not establish performance for users in another region.
- One successful ping does not prove that a TCP service, DNS resolver or long-lived transfer is healthy.
Prepare an escalation package for the provider
Support can act faster when the report is reproducible. Include:
- service or instance ID, affected public IP, operating system and VPS region;
- the affected destinations, ports and protocols;
- UTC timestamps, client location and client network;
ip addr,ip routeand relevant firewall output;- MTR or traceroute reports, including the TCP command used for TCP services;
- iperf3 results in normal, reverse, multi-stream and, where relevant, bidirectional modes;
- CPU, memory, swap, disk-latency, I/O-wait and virtual-NIC graphs covering the test window;
- the expected behavior, a normal comparison destination and the exact time the problem began.
State whether loss continues to the destination, whether both directions are affected, and whether the issue follows a region, protocol or time of day. Keep console access available while support investigates routing, IPv6, multi-network packet loss or a possible MTU issue.
A practical decision path
- Classify: latency, loss, DNS, application delay or throughput; note direction and destination.
- Validate the guest: addresses, default route, firewall and listening socket.
- Test naming: compare direct IP behavior with hostname resolution and a real service request.
- Trace the path: use MTR or traceroute, matching the protocol and port where possible.
- Measure capacity: run controlled iperf3 tests in both directions, with single and multiple streams.
- Correlate resources: check CPU, memory, swap, disk, NIC and provider graphs during the test.
- Verify entitlement: compare results with the exact VM size, region, egress and shared-port terms.
- Test MTU cautiously: only when packet-size and protocol evidence points there, with console rollback access.
- Escalate with evidence: send commands, outputs, timestamps and graphs rather than the word “slow” alone.
Frequently Asked Questions
Can a VPS be network-limited even when its CPU is mostly idle?
Yes. A virtual-port or provider egress ceiling, vNIC saturation, host contention, disk I/O interference or a congested route can limit throughput without high guest CPU usage.
Best Value
- 5-in-1 Connectivity: Equipped with a 4K HDMI port, a 5 Gbps USB-C data port, two 5 Gbps USB-A ports, and a USB C 100W PD-IN port. Note: The USB C 100W PD-IN port supports only charging and does not support data transfer devices such as headphones or speakers.
- Powerful Pass-Through Charging: Supports up to 85W pass-through charging so you can power up your laptop while you use the hub. Note: Pass-through charging requires a charger (not included). Note: To achieve full power for iPad, we recommend using a 45W wall charger.
- Transfer Files in Seconds: Move files to and from your laptop at speeds of up to 5 Gbps via the USB-C and USB-A data ports. Note: The USB C 5Gbps Data port does not support video output.
- HD Display: Connect to the HDMI port to stream or mirror content to an external monitor in resolutions of up to 4K@30Hz. Note: The USB-C ports do not support video output.
- What You Get: Anker 332 USB-C Hub (5-in-1), welcome guide, our worry-free 18-month warranty, and friendly customer service.
Should I use ping, traceroute or MTR for an HTTPS problem?
Use all three in sequence, but include a TCP MTR to port 443, such as mtr DESTINATION -T -P 443 -c 100 --report. Judge a hop by loss that persists to the destination, not by an isolated non-response.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →When is changing the VPS MTU justified?
Only when large-packet or UDP failures, fragmentation or path-specific stalls point to an MTU mismatch. Make the change from console access, save the original value and keep a rollback plan.
How do I test upload and download separately?
Run iperf3 normally for traffic from the VPS to the test server, then repeat with -R for traffic from the server to the VPS. Compare single-stream, multi-stream and, if the workload is simultaneous, bidirectional results.
Why is SSH slow when a file transfer is fast?
SSH interactivity is sensitive to latency, jitter and packet loss, while a bulk transfer can maintain throughput despite delay. Run ping and a TCP MTR to the SSH destination, and check for route changes or sustained loss.
What information should I give VPS support?
Provide the instance ID, public IP, OS, region, destinations and ports, UTC timestamps, client location, iperf3 results in both directions, MTR or traceroute reports, route and firewall output, and resource graphs from the test window.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsThe Bottom Line
Diagnose a slow VPS by separating DNS, latency, loss, application delay and throughput. Verify the guest, trace the real path, measure both directions with controlled iperf3 tests, correlate resource and provider limits, and change MTU or firewall settings only with console access and a rollback plan.
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.
