Thursday, 27 February 2025

why speed test results are usually lower than advertised speed

Overhead

  • Your internet speed is measured in raw bits per second (bps), but real-world data transfer includes additional information like headers, acknowledgments, and error checking.

  • TCP/IP, Ethernet, and other protocols add packet overhead, meaning some portion of the bandwidth is used for network management rather than your actual data.
  • This overhead typically accounts for 5-15% of the total bandwidth, which explains why you rarely see a full 1 Gbps in speed tests.

Speed test client and server

  • The speed test server’s capacity can impact results. Some servers may be congested or unable to fully utilize your bandwidth.
  • The distance between you and the test server affects latency, which can slightly reduce speeds.
  • The speed test client needs to have a good NIC (intel if possible) and good spec (RAM and CPU) because it needs to make many connections to test the connection. For example if you test with a 100mbps NIC that is the max speed you can see. You need a 1gig NIC or better to test a 1gig connection.

  • ISP and firewall/network management and contention

    • ISPs often use network shaping, congestion control, and peering agreements that affect speed.
    • During peak times, ISPs may limit speeds slightly to ensure fair distribution of bandwidth among users.
    • Your internal network/firewall may do the same
    • Its best to test out of hours with just your test laptop plugged into the internet connection to give the best results

    Thursday, 20 February 2025

    check a DNS TXT record

     

    https://mxtoolbox.com/SuperTool.aspx?action=txt%3a%40.dlrcoco.ie&run=toolpage#

    nslookup -type=TXT mail._domainkey.domain.com


    dig TXT domain.com +short


    Powershell

    Resolve-DnsName -Type TXT domain.com

    Wednesday, 12 February 2025

    Meraki monitoring pages

    Organisation -> Alerts

    Network-wide > clients

    Network wide > Traffic