Kitz ADSL Broadband Information
adsl spacer  
Support this site
Home Broadband ISPs Tech Routers Wiki Forum
 
     
   Compare ISP   Rate your ISP
   Glossary   Glossary
 
Please login or register.

Login with username, password and session length
Advanced search  

News:

Author Topic: Need Help - Latency & Packet Loss  (Read 603 times)

neil

  • Reg Member
  • ***
  • Posts: 502
Need Help - Latency & Packet Loss
« on: March 28, 2020, 06:30:01 PM »

I launched complaint with my ISP for packet loss and slow speeds during evening and night time few days ago. Today i received a reply that complaint is solved but it is not. I sent pingplotters screenshot and speed test files. Can you guys help what kind of test i need to do to convince my problem is not solved yet.
I am attaching ping plotter and speed test text file.
First one is with pkt loss.
second one is speed test graph.
third one is ping plotter without pkt loss.
Logged
VDSL FTTC 35/18

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Need Help - Latency & Packet Loss
« Reply #1 on: March 29, 2020, 04:11:30 AM »

Could also take a look at

    https://www.thinkbroadband.com/broadband/monitoring/quality and
    http://fruk.net/index.php?fruk=f8lure

which use a Firebrick ping server. They are rather different.

Could perhaps set up an iPerf performance measurement session with it set to use UDP instead of the usual TCP, but be very careful how you do this and read the manual first as you will break the internet if using UDP mode if you’re not very careful and your ISO may well see this as a DOS attack. I made this mistake. You would have to carefully tune the speed to the max your internet connection can handle and then back it down a bit. Obviously you would get normal packet loss when you go to fast and overload your link. Some packet loss is normal under certain conditions when a link is genuinely overloaded by traffic and indeed it’s even required, as eg TCP requires packet loss amongst other things to signal it to slow down. That together with timing measurement and maybe ECN if available. Ask for some help if you want to try an iPerf experiment.

All these have a problem the servers I mentioned are based in the wrong country, so you could end up getting congestion in the far off reaches of the internet so you’re then measuring the wrong thing, not just testing your own line. Ideally you need a test server parked at your ISP.
Logged