Azerbaijan, Georgia, Uzbekistan
XcitiumRequest a demo
MenuClose
All materials

Network foundations · Protocols

TCP, ping and TTL: understanding a network connection

TCP supports reliable delivery of a data stream, ICMP carries diagnostic messages, and IP TTL prevents packets from circulating indefinitely. This guide links those concepts to network topology and simple read-only checks.

Four server models connected to a central network node

Key takeaway

Ping and traceroute offer clues about a path, but one measurement cannot establish the health of an entire network or application.

Topology describes the arrangement

Topology is the physical or logical arrangement of devices and links. In a star, devices connect to a central switch; a mesh offers alternate paths; tree and hybrid designs combine patterns. A bus shares one central link, while a ring connects devices in a loop; these forms are less common in modern LANs. The physical cable map can differ from logical VLAN and subnet boundaries. The impact of a failure depends on the design.

Star

Devices connect to a central switch; if it fails, those connections are affected.

Mesh

Multiple paths can add resilience while making routing and management more complex.

Logical segment

VLANs and subnets group devices by policy independently of physical cabling.

IP packets, TCP and UDP

IP packets carry source and destination addresses, and routers forward them toward another network. TCP establishes a connection with a SYN, SYN-ACK and ACK exchange and can retransmit lost data to provide an ordered byte stream. UDP offers a lighter datagram service without delivery or ordering guarantees. A SYN flood can target TCP resources; TCP’s reliable delivery does not encrypt data, so confidentiality requires a separate protocol such as TLS. HTTP/HTTPS carries web content, DNS resolves names, and DHCP supplies network configuration.

What ping checks

Ping sends an ICMP Echo Request and reports round-trip time (RTT) when a reply arrives. A limited set of requests can check whether a host or gateway responds. The result describes only the ICMP path at that moment: no reply does not prove the host is off, and high latency alone does not prove an application is slow because intermediate networks may treat ICMP differently. Large volumes of ping traffic can be abused to disrupt a service or identify active devices, so keep diagnostics limited and authorized.

      ping -n 4 example.com
    

IP TTL and DNS TTL are different values

The TTL field in an IPv4 packet decreases as it passes through routers; when it reaches zero, the packet is discarded and an ICMP Time Exceeded message may be sent. IPv6 uses a field named Hop Limit for the same hop-limiting purpose. Traceroute raises the limit step by step to collect replies from intermediate routers. A DNS record’s TTL states how many seconds a resolver may cache the answer; application and session caches can have their own TTLs.

Interpret missing replies in a route trace

On Windows, tracert /d example.com increases TTL to collect replies along a path; /d skips name lookups for intermediate addresses. An asterisk can mean a router did not reply, but it does not establish that the route is broken at that hop. Compare the result with a repeat check to the same destination, name resolution and the application’s own availability test.

      tracert /d example.com
nslookup example.com
    

A simple diagnostic sequence

To narrow a problem, check the adapter and gateway first, then reachability by IP, name resolution and finally the application. Ping and traceroute provide network-layer clues but do not identify a fault on their own. In an organization, inspect logs, routes and firewall rules only with authorization; avoid scanning address ranges or sending continuous probes.