I am using OSMC + Transmission. Whenever transmission is active and something is downloading it consume all my bandwidth. Everything works fine when I stop the service.
How I can limit traffic? Which are the correct values for download and upload bandwidth?
I use a 300 Bps wifi router Netgear N300 where I usually connect 2 phones, 1 tablet, 1 PC and raspberry.
Considering sometime I download .avi, .mkv or pdf files (PDF usually are opened on PC) what are the security risks? .avi and .mkv are played on raspberry. Is it possible my network is infected by Transmission?
I’ve installed tranmission from appstore and i got the web interface at “my ip”:9091, in the lower left corner there is a wrench icon. That gives a little window with options, there is a tab called speed, where you can set upload and download limits.
Well opening a port is crucial if you wanna share back what you have downloaded, more ppl will be able to connect to your torrent client and download from you. And some sites require you to be “connected”, that is have an open port direct to your torrent client.
When it comes to speed i’ve never needed to limit my download or upload, pi connected via cable to router. And i’m on a 100mbits internet connection. You haven’t mentioned your internet connection speed, so i’m saying this in hope that you atleast got 10mbits upload and download speed. If your WiFi/internet gets hogged up by the torrents running, try set your max it at 5000 on both upload and download. This should run okey on a 10/10 connection.
Transmission usually grabs as much bandwidth as it can unless you limit it. Start by running it consuming all the bandwidth, no limit. Look at the speed it’s downloading at, and then set a limit of about half the maximum speed. If the rest of the system (osmc etc) still seems slow, reduce the limit further. If the system is running ok with transmission using half the bandwith, increase it until you get a value you are happy with.
Another option is to use the Alternate Speed Limit settings in transmission so that for example you can run at full speed overnight when nobody is using your internet connection much, and slug the speed to half or less in the evening when you want to be able to do other things simultaneously
Ok, thank you everybody. I’ll try your suggestions.
However, I do not use Ethernet cable. All devices are connected via WIFI.
I have a Netgear router DNG2200 with 300 Mbs.
I am doing some tests right now because I want understand performance of my network before apply limits to transmission.
Doing these tests I discovered that Router-Raspberry connection speed was 10 times slower than PC-router.
TEST 1:
I am transferring a 1.2 Gb file from a PC to Raspberry both connected via WIFI (the result was 5 Mb/s too slow).
TEST 2:
Connecting PC via ethernet 10Gbit the transfer speed increased only to 6 Mb/s.
TEST 3:
Connecting Raspberry via ethernet 10Gbit the speed increased to 61 Mb/s.
Then I find out that my Raspberry was in a very bad position, in fact put it close to WIFI router the TEST 2 speed increased to 30 Mb/s
The question is: why Raspberry-Router connection is slower (30 Mbs) than Router-PC connection (60 Mbs)?
Consider that on my Raspberry I use a WIFI 820n dongle (the OSMC official one).
I know wifi dongle are connected using USB 2 that has a 480 Mb/s limit -> 60 MB/s.
However, considering all these tests I think before work on transmission settings I need to understand if my wireless connection works properly. From value of TEST 3 I think it is too slow. I expected to find something around 140 Mb/s instead of 61 Mb/s.
Consider all tests have been done also putting PC and Raspberry close to the Router. I also rebooted the router to be sure there is no overlap with other wireless networks.
What do you think about these tests? What can I do to increase my wireless network speed?
I have only one doubt. Few months ago my kid started to play with one of router Antenna. He removed the plastic of one antenna (but the wire is still there) and I wasn’t able to atacch it again because the plastic was broken. Now at that time I saw everything worked properly and I left everything as it was. I think the antenna wire is still connected but not totally sure. I do not know what could be the impact of this on router performance.
That router technically only Runs at 150Mbit for Wireless N connections. 300mbit is actually the combination of all wireless modes combined, while the maximum throughput for strictly Wireless N is only 150mbit. It’s just marketing.
You might want to log in to the router and make sure that the router mode is set for 150mbit (Wireless N+G)
If you have it in compatibility mode (WirelessB+A) it will be restricted to 65mbit.
Keep in mind that the 150mbit limit is the combined total bandwidth available to the wireless router. The more wireless devices you add, the more that gets carved up, so don’t expect to have 150mbit to all devices.
I have a much higher end router than that at home (Asus RT-N66U, N900 as opposed to N300) and I’m lucky to get 100mbit 15 feet away.
I would expect that your router is actually behaving as it is supposed to (It’s not a super high end router)
Well there are many things that can impact wireless throughput and e.g. Getting 15M throughput on a 54M connection is totally possible based on congestion and interference.
Suggest to first get proper test tools like iperf3, wavemon and e.g. an Android phone with Wireless Analyser to check channel congestion.
Also certain channels have certain restrictions in certain countries which can slow down things based on configured country
Let me understand. I read in the past stuff like this but I am not an expert. For the moment I only know that my router was configure to 54 Mbs so I accessed to it and changed it to 300 Mbs. I saw there is also the option to 150 Mbs. The question is: why I should reduce speed to 150 Mbs? Why reducing the speed should convert in an increase of throughput? That’s in counter intuitive.
How I can check the compatibility mode? What is Wireless A and B?
Consider that all tests have been done with devices connected close to router. Moreover, only my PC, Raspberry and Router were turned on, all the others devices (tablet, phones) were turned off.
I also checked with some apps if my router channel overlap some other wifi networks.
I know but as I said I did the test reducing all the possible variables.
I haven’t used iperf3 but I think a copy of a large file measuring the time is perfectly equivalent. Am I correct?
You also have the USB 2.0 speed limitations for your wifi dongle. That should cap the bandwith at 60MB/s.
So if you get 30MB/s you are already very lucky on wifi as the dongle will not be the only thing on that bus and not the only wifi device on your home network, in addition to signal loss in wireless land in general.
The ethernet chip is also connected via USB, but you seem to be using it fully there.
You’re right. I expected the Rasperry-Router link speed was 60 Mb/s but I found the half (30 Mb/s). Raspberry was very close to WIFI router. USB is 480 Mb/s → 60 MB/s so I expected to find this cap. This morning (Rome time) I thought two things:
on Raspberry USB was connected also the HDD (that was not used during the transfer) I do not know if there was an impact, I can repeat the test without HDD.
I stored the file on Raspberry SD card and I just verified it is a class 10 (10 MB/s) so this could be the reason of the cap. The only thing I do not understand is that 30 Mb/s is almost 4 MB/s that is much lower than 10 MB/s.
I tried also to store the file on mounted HDD (connected to Raspberry via USB) but the speed was very very slow. The same file took almost 40 minutes to transfer. Why HDD connected via USB is slower than a SD Card?
For USB you will not see a cap at 60MB/s. Around 30-45 will be much more likely.
If you transfer files over network to a usb harddrive, you will basically get half the bandwidth as it is shared on the same usb bus.
Hi,
Going back to my original problem I need to limit Transission traffic so that other device should not have impact during navigation. I read this tutorial:
and I have the parameters to put in transission. My question is why it take in consideration only upload speed? Why it suggest me 0 for download speed?
I ask this because yesterday I tried several values without success, so I think I have to understand better how things works.
If you run speedtest.net and chose the nearest server to you, you should get an accurate upload speed messured. From that speed you can calculate how much you want to put in upload limitation on your torrent client.
for example i got: 11,41 mb/s and if i wanted to limit my upload speed i’d limit it to 9000kb/s so that i have upload room for other devices on my network. Thats depending on outgoing traffic alone.
I haven’t taken in any other variables on my pi/transmission setup, i download to the hard drive and use ethernet so they are sharing the bus, but i seldom get faster download or upload speeds so the USB-bus gets overloaded. I don’t think that will your problem either, your main concern should be upload speed of your connection, second your wifi. So in general take your upload speed and cap your torrent client at 2/3’s of your upload speed, if that still make your wifi congested then lower the limit till you get good wifi on your other devices.
Hi,
I just performed a test with server in Rome and I live in Rome too. Here the result:
6.86 Mbs in download
0.42 Mbs in upload → 420 Kbs
Italy adsl connections are not so fast. Fast networks cover only several zone in Rome.
Now my first question is why I should care about upload speed? Why I should not consider download speed that make more sense.
2/3 of 420 Kbs is 280 Kbs.
Now where I have to put this value? Transmission has several parameters. Is this a cap for upload or download speed? If this is for upload, what about download?
Above I posted this tool http://infinite-source.de/az/az-calc.html
that from upload speed returns all the others parameters. The thing I do not understand is why it is always set download cap to 0?
In general it seems that upload speed is the discriminant to calculate the cap (you said the same in your post) but I do not understand, from a theorical point of view, why.