文章
网络技术的历史与演变:以太网、WiFi 和 VPN。
Network Troubleshooting: A Step-by-Step Diagnostic Method
When the network breaks, work in a logical order. Learn how to use ping, traceroute, DNS lookups and port checks to isolate a problem, and what each tool actually reveals.
How DNS Works: A Practical Guide to Domain Name Resolution
Understand how DNS turns a domain name into an IP address, the difference between recursive and authoritative servers, how caching and TTL work, and how to diagnose a resolution problem with a lookup tool.
Port Forwarding vs Port Scan: What Is the Real Difference?
Opening a port on your router is not the same as testing whether it is reachable. Learn the difference between port forwarding and port scanning, why people confuse them, and how to verify that a forwarding actually works.
IPv4 vs IPv6: Address Exhaustion, NAT and How to Diagnose Both
Two versions of the IP protocol run in parallel today. This guide explains how their addresses differ, why IPv4 ran out, what NAT does about it, and how to diagnose both on your network.
How a Speed Test Works: Download, Upload and Why Results Vary
The number from a speed test is the product of a short measurement against a remote server, not a fixed property of your connection. This guide explains what it measures, why it fluctuates, and how to get a reliable result.
SSL/TLS Certificates Explained: Trust Chains and the Most Common Errors
Certificates are what turn the padlock icon from a warning into a reassurance. This guide explains the chain of trust, how to verify a certificate, and what the common HTTPS errors actually mean.
WHOIS Explained: What a Domain Lookup Reveals
A WHOIS lookup shows the public record behind a domain: registrar, dates, nameservers and status. Learn what it reveals and what privacy often hides.
DNS Propagation Explained: Why Your DNS Change Takes Time
DNS changes do not appear instantly everywhere. Learn how TTL and caching servers control propagation, and how to check a change from multiple resolvers.
Subnetting Explained: How IP Subnet Masks Divide a Network
Learn how a subnet mask splits an IP network into smaller, manageable subnets - and how CIDR notation, host counts and supernetting all fit together.
Public vs Private IP Addresses: What Your IP Reveals and Why It Matters
Every device on a network has an IP, but only one kind is visible to the world. Understanding public versus private addressing explains what your IP reveals, what it hides, and why it changes.
How to Read a Traceroute: Every Hop, Every Asterisk, and When a Route Is Abnormal
Traceroute shows the path your packets take through the internet. Reading each hop - the latency, the asterisks, the private address - turns a route into a real diagnosis of where a problem lives.
How to Read a Port Scan: Open, Closed and Filtered Ports Explained
A port scan returns a wall of open, closed and filtered ports. Knowing what each state actually means turns a raw list into a real answer about what is exposed and what is safe.
DNS Record Types Explained: A, AAAA, CNAME, MX, TXT, NS and PTR
DNS is not one lookup but many, distinguished by record type. Knowing what each type answers turns 'the domain doesn't work' into a two-minute diagnosis - and explains half of what mail and verification tools do.
The Four HTTP Security Headers Every Site Should Send
HSTS, Content-Security-Policy, X-Frame-Options and X-Content-Type-Options cost an afternoon to deploy and quietly prevent whole classes of attacks. Here is what each one does and how to adopt them without breaking your site.
Blacklists (DNSBL): Why Your IP Gets Listed and How to Get Off
Mail servers check every sender against public blocklists. A single listing can silently drop your emails - here is how the system works, why innocent IPs end up on it, and what actually gets you delisted.
Ping vs Traceroute: Which One Answers Your Question?
Both tools test reachability, but they answer two different questions. Knowing which one to reach for turns a vague 'the site is slow' into a precise diagnosis in seconds.
Why These Tools Exist: Reading Results, Judging Warnings, and When You Actually Need a VPN
What each diagnostic tool is actually for, how to read a result without over- or under-reacting, concrete open-port risk examples, and when exposure genuinely calls for a VPN or firewall change - versus when it's normal.
The Rise of the VPN: From Dial-Up Tunnels to WireGuard and Zero Trust
How VPNs evolved from a niche way to reach the corporate LAN into the default way anyone protects traffic on an untrusted network - and what each generation of protocol actually changed.
The Genesis of WiFi: From 2 Mbps to Multi-Gigabit, Generation by Generation
How WiFi went from a slow, unreliable 1997 curiosity to the multi-gigabit, multi-device backbone of every home and office network today.
From LAN to the Internet: Token Ring, Ethernet, and the Protocols That Won
How local networks went from proprietary token-passing rings to the switched Ethernet and TCP/IP stack that now underpins the entire Internet.
我的 IP 检测
自动检测您的公网 IP 地址和网络位置。
打开 NetChecks 时自动加载，无需输入。切换网络或重新连接 VPN 后，点击刷新按钮重新检测。
您的浏览器
一键全面扫描
对一个 IP 或主机名一次性运行所有相关检测:DNS、whois、ping、traceroute、知名端口扫描 (1-1024)、HTTP 头信息和 SSL 证书。
输入域名或 IP 地址并运行，一次性检查 DNS、whois、ping、traceroute、常用端口、HTTP 响应头和 SSL 证书。
大多数检查并行执行,通常在约30秒内完成,若目标响应缓慢或无法访问则更久。
只有在上方的同意复选框被勾选后，端口扫描步骤才会执行——其他所有检查照常执行。
Ping
向主机发送 ICMP 回显请求,以检测其可达性和延迟。
输入主机名或 IP 地址，点击 Ping 发送 ICMP 回显请求并测量往返延迟。
深入了解 Ping
定义
Ping通过向目标主机发送ICMP回显请求(Echo Request)报文,并测量ICMP回显应答(Echo Reply)返回所需的时间,来完成最基础的网络连通性测试。它只回答一个问题:"这台设备是否可达,响应速度如何?"ICMP协议(RFC 792)早在1981年就被设计用来在IP网络上承载控制与诊断信息,独立于应用层流量之外——Ping正是它最广为人知、也是普及程度最高的实现,几乎所有操作系统和网络设备从最初版本起就自带这一功能。
工作原理
每个ICMP报文都携带一个TTL(生存时间)字段,每经过一台路由器就减1;一旦在到达目标之前减到0,报文就会被丢弃,并向发送方回送一条错误消息。以毫秒为单位测得的往返时延(RTT)反映的是整个往返路径上累积的网络延迟,而不仅仅是目标附近最后一段的延迟——这一点常被误解,因为Ping结果慢的根本原因可能出现在路径上的任意位置,未必靠近被测服务器。Ping通常会连续发送多个报文而不是只发一个,这样就能区分出偶发的延迟毛刺和持续性的问题,并据此计算出这一批样本的丢包率。
结果解读
稳定且较低的RTT——局域网内几毫秒,同一国家内的目的地10到50毫秒,跨洲链路则明显更高——说明连接状况良好。哪怕只是轻微的丢包(超过1%到2%),对VoIP或交互式远程会话这类对延迟敏感的场景也会造成明显影响,每一个丢失的包都会表现为卡顿或断音。相比稳定但数值较高的延迟,包与包之间延迟波动很大(抖动)对这些场景往往是更严重的问题。"请求超时"表示在规定时间内没有收到应答——可能是主机确实宕机,也可能是防火墙悄悄屏蔽了ICMP,或者路径上某处出现了故障;"目标不可达"则不同,信息量更大:它是路径上某台中间路由器主动回送的消息,明确表示自己无法转发该报文,有助于缩小问题定位范围。
常见误区
最常见的误判是一旦Ping失败就断定主机"宕机",而实际上大量服务器和设备——尤其是部署了严格防火墙策略或托管在主流云服务商那里的——出于策略考虑会主动屏蔽入站ICMP,同时它们提供的实际服务(HTTP、数据库等)完全正常可用。因此,没有收到Ping应答只有在结合其他信号(比如应用本身也无响应)时才具有实际意义。反过来,Ping成功也完全不能保证该主机上运行的应用服务本身工作正常——这是网络栈中两个完全独立的层面。
适用场景
在升级工单之前先做的第一项检查:先确认设备是否有响应,再深入排查。在修改防火墙规则或路由表之后确认连通性,确保改动没有破坏访问。在VoIP上线或运营商链路切换之前建立延迟基线,以便后续出现通话质量投诉时有客观的对比依据。作为MSP并行监控多个客户站点时一种轻量、低开销的周期性健康检查手段,但始终应作为更深层应用级监控的补充,而非替代。
Traceroute(路由追踪)
逐跳追踪到目标主机的网络路径。
输入主机名或 IP 地址并运行，查看本服务器到目标之间的每一跳，以及各跳的延迟。
DNS 查询(Nslookup)
查询 DNS 记录:A、AAAA、MX、TXT、NS、CNAME、SOA、PTR、SRV、CAA。
输入域名，选择记录类型（A、AAAA、MX、TXT、NS、CNAME、SOA、PTR、SRV 或 CAA），然后查询。
Whois 查询
查询域名或 IP 地址的注册信息。
输入域名或 IP 地址，查询其注册信息：注册商、所有者组织及重要日期。
黑名单检测
检查某个 IP 地址或域名是否被列入公共垃圾邮件/滥用黑名单(DNSBL)。
输入 IPv4 地址或域名并运行,即可一次查询 7 个公共 DNSBL/RBL 黑名单 - 每个列表会显示已列入、未列入或检测失败。
TCP 端口扫描
检测主机或 IP 上的 TCP 端口是否开放:常用端口、自定义列表,或完整的 1-65535 范围。
输入主机或 IP，选择常用端口、自定义列表或完整范围，然后扫描查看哪些 TCP 端口有响应。
完整范围扫描最长可能耗时约 100 秒。大范围扫描(超过 100 个端口,包括自定义范围/列表)每位访客每分钟限 1 次。
开放端口 (0)
HTTP 响应头检测
获取某个 URL 的 HTTP 响应状态码和响应头。
输入网址，获取其 HTTP 响应状态码以及服务器返回的所有响应头。
SSL / TLS 证书检测
检测主机的 TLS 证书:颁发机构、有效期以及剩余天数。
输入主机名，检查其 TLS 证书：颁发者、有效期及距过期的剩余天数。
IP 地理位置查询
查询某个 IP 地址的地理位置和网络信息。留空则查询您自己的公网 IP。
输入任意 IP 地址，或留空以查询您自己的 IP，查看其大致位置及网络/ISP 信息。
子网 / CIDR 计算器
完全在您的浏览器中计算 — 不会向服务器发送任何数据。
输入 IP 地址和 CIDR 前缀（例如 192.168.1.0/24），即时计算网络范围、广播地址和可用主机数量。
网速测试
与本服务器之间的基础下载/上传速度测试(准确性取决于服务器自身的网络带宽)。
点击开始，测量与本服务器之间的下载和上传速度。准确度取决于本服务器自身的网络连接。
国家代码词典
ISO 3166-1 alpha-2 国家代码 — 完全在您的浏览器中搜索。
搜索或浏览 ISO 3166-1 alpha-2 国家代码列表，完全在您的浏览器中查询。
| 国家 | ISO 代码 |
|---|
国际电话区号词典
按国家划分的国际电话区号 — 完全在您的浏览器中搜索。
按国家搜索或浏览国际拨号代码，完全在您的浏览器中查询。
| 国家 | 区号 |
|---|
世界时钟
选择一个时区查看当前时间 — 拖动地球即可旋转。
从列表中选择一个时区，或拖动地球仪，查看该地的当前时间。
—
—
拖动以旋转地球。
法国移动网络状态
法国各运营商（Orange、Free、SFR、Bouygues Telecom）故障或维护中的移动基站，数据来自 Arcep 公开数据。每日更新一次快照，非实时数据流。
按法国运营商浏览移动基站和光纤故障数据——无需输入，根据 ARCEP 公开数据自动更新。
受影响最多的省份
当前故障或维护中的站点数量，按省份统计。点击上方运营商可筛选。
数据来源： Arcep — Sites indisponibles · 官方网络状态地图
固网（光纤）
各运营商光纤（FTTH）网络质量：故障上报率与装机失败率，数据来自 Arcep 公开数据。月度指标，6个月滚动平均——不像移动部分那样为实时数据。
按运营商（母公司）
过去 6 个可用月份的平均值，按基础设施运营商母公司分组。