Large Internet Speed Test Download

[rpiswag]

How to do a 5 gb speed test download in terminal. I know it is large but I still want to do it. I want to save it to dev/null so it will be automatically deleted. I have many pis and I want to test my network so how can I have the file be re downloaded over and over? I want to put my 100megabit per second connection to it's limits.

[aTao]

This site may be of use. http://speedtest.tele2.net/

[rpiswag]

Thank you so much! How can I get the 1 gb file download test to automatically download over and over?

[kusti8]

Just make an infinite loop:

Code: Select all

while true; do (code); sleep 2; done

[rpiswag]

Does the file automatically get deleted and where do I put the code you posted and where is the file the code goes into? Can you post the edited code with your edits so it will download over and over. What I am saying in a nutshell is to post the entire code and step by step instructions.

[kusti8]

Simply input this into the terminal and it will infinitely download the 1GB file. It will be deleted automatically as well.

Code: Select all

while true; do wget -O /dev/null http://speedtest.tele2.net/1GB.zip; sleep 2; done 

Edit: you can quit it with CTRL+C. It gets output to /dev/null, or the trash in so nothing is saved. You put this in the terminal.

[rpiswag]

With that website my download speed dropped very low so I decided to use wget --output-document=/dev/null http://speedtest.wdc01.softlayer.com/do ... st1000.zip but with it in a infinite loop with the command you showed me and I am getting much faster speeds. Why am I getting WAY faster speeds?

[gkreidl]

With that website my download speed dropped very low so I decided to use wget --output-document=/dev/null http://speedtest.wdc01.softlayer.com/do ... st1000.zip but with it in a infinite loop with the command you showed me and I am getting much faster speeds. Why am I getting WAY faster speeds?

Did you try it with a browser? Webkit browsers by default load it into memory first. Your RPi might have started swapping.

[rpiswag]

Can I do the download test for months on a 4 gb sd card and will I use up my memory with these tests?

[gkreidl]

Can I do the download test for months on a 4 gb sd card and will I use up my memory with these tests?

Why? Don't you think about the traffic you're imposing on the provider of the test file? Traffic costs money and somebody always has to pay for it. I suppose sooner or later the site will block you.

[Heater]

rpiswag,

Can I do the download test for months on a 4 gb sd card and will I use up my memory with these tests?

What you are proposing to do is generally referred to as a Denial of Service Attack, DOS attack. This generally seen as a bad thing. People go to jail for doing this. I have no idea where you are but in the UK:

The Police and Justice Act 2006, of the United Kingdom, amended the Computer Misuse Act 1990 and specifically outlawed denial-of-service attacks and set a maximum penalty of 10 years in prison.

Or in the USA:


...denial-of-service attacks may be considered a federal crime under the Computer Fraud and Abuse Act with penalties that include years of imprisonment.


Other countries have similar laws.

What with your ongoing password cracking project and a Linux install full of hacker tools your future is not going to be good when the cops come a knocking...

[aTao]

With that website my download speed dropped very low so I decided to use wget --output-document=/dev/null http://speedtest.wdc01.softlayer.com/do ... st1000.zip but with it in a infinite loop with the command you showed me and I am getting much faster speeds. Why am I getting WAY faster speeds?

The slow down is due to the SD card, writing very large files can slow to a crawl. Writing to /dev/null dos not go anywhere near the SD card, in fact it just evaporates.

You really should read the @fair use@ policy on the source web sites too, they ask quite nicely that you do not hammer them as you propose.

What is the point of your test anyway, continuously monitoring network speed for 4 months is a b it strange

[rpiswag]

Thats if you do it to some one else. I am having my cousins over and they use all of the internet connection and I am trying to simulate it. I found a awesome setting in my router that will give a temporary speed boost to each device so it can load faster. Currently when I ran the test on some the internet connection almost never dropped. I expected it to drop 5-10 mbps less but that still leaves me with another 45mbps. Whey don't the pi use the entire network connection. This is why I did this because I expected one thing and got another. Don't the servers who host the speed test have 10gbps internet connections? I tested it again and it dropped about 5mbps but when I get my computer closer to the speedtest gets to 50mbps. The Raspberry Pi is connected through wifi(IEEE802.11).

[DougieLawson]

Wouldn't it be better to set up your Raspberry Pi as a captive portal or proxy and force your cousins to run their traffic through your system? You can then learn about captive portals, local proxies, throttling and traffic shaping and other useful networking systems.

What you're proposing is complete madness and a complete waste of bandwidth at both ends of the connection.

[aTao]

Thats if you do it to some one else. I am having my cousins over and they use all of the internet connection and I am trying to simulate it. I found a awesome setting in my router that will give a temporary speed boost to each device so it can load faster. Currently when I ran the test on some the internet connection almost never dropped. I expected it to drop 5-10 mbps less but that still leaves me with another 45mbps. Whey don't the pi use the entire network connection. This is why I did this because I expected one thing and got another. Don't the servers who host the speed test have 10gbps internet connections? I tested it again and it dropped about 5mbps but when I get my computer closer to the speedtest gets to 50mbps. The Raspberry Pi is connected through wifi(IEEE802.11).

Sure the servers have a fast connection, but they 1 pay for it and 2 expect many people to be able to use the service. The idea is for occasional testing and trouble-shooting.

If you just want to hammer your connection it would be better to get as many streaming video feeds running on your network (TV sites preferably). The site I mentioned is more aimed at people that need to spot and understand any traffic shaping that is being applied to them, thats why they offer the same media as http and ftp and various other protocols.

[davidcoton]

Thats if you do it to some one else. I am having my cousins over and they use all of the internet connection and I am trying to simulate it. I found a awesome setting in my router that will give a temporary speed boost to each device so it can load faster. Currently when I ran the test on some the internet connection almost never dropped. I expected it to drop 5-10 mbps less but that still leaves me with another 45mbps. Whey don't the pi use the entire network connection. This is why I did this because I expected one thing and got another. Don't the servers who host the speed test have 10gbps internet connections? I tested it again and it dropped about 5mbps but when I get my computer closer to the speedtest gets to 50mbps. The Raspberry Pi is connected through wifi(IEEE802.11).

So do you know the bandwidth of each part of the path? Which is the weakest link in your test? When your cousins visit, are they using a path with the same weakest link? Is the test even simulating the right conditions? And does the short test tell you anything unexpected? What are you going to do about it anyway?

For the next bit I won't ask, I'll just tell you: A three month long speed test will not tell you anything useful that a standard short speed test won't tell you. (Why would it?)

I admire your energy and persistence. But if you are going to be a useful engineer, you need to learn to plan your projects. What is your goal? How do you hope to achieve it? What are the stages towards solution, each stage to be a "gateway" that must be achieved before it is worth setting up the next stage? What equipment is needed for each stage? What investment is required? What is the payback (not necessarily just in financial terms)? When do you stop and abandon a project before wasting too much time and energy? At present you seem to be firing off in all directions, with multiple projects, some extremely ill-advised, and most relying on asking what to do -- so you discover nothing original, and can only repeat what others tell you they have already done. Try to find projects that you can research yourself, that you can realistically try out (ie, they are only one or two steps ahead of what you already know), and that you can then use to carry on to the next level. Very soon you will find you are doing original research and finding useful new ways of using computers. Then you might even develop the Next Big Thing (TM).

Look back at your posts here. People have asked you many questions to guide you, but too often you have not responded and gone off in a different direction. People will soon get fed up and will stop offering advice if you continue that way. Take the advice, try what is suggested, and report back what you have learnt. Then people will know you are making progress, and will want to continue to help.