I have two routers connected to each other via a CAT6 cable.
Router A is 1100AH, Router B is a 450G. The link between the two is a full-duplex 1Gbps connection.
How come I can only get ~60Mbps of full-duplex speed on the bandwidth test?
Try using “UDP” tcp will be slower due to the talk backs.
If I only test Receive on TCP, it can only get to 150Mbps. Whenever I have tested bandwidth on a 100mbps connection, I usually get about 90-95mbps. How come I can’t get to gigabit speeds(or close at 900mbps) on Gigabit interfaces?
Nobody said that this Hardware is capable to run at this speed. It has just Gbit interface, but it is not able to handle data at this speed.
Cpu is on 100%. Do have not enough power to Btest in TCP
I guess I was expecting more out of the RB1100AH and Gigabit Interfaces.
You are doing the test wrong. Btest generates random data, so to generate this random data for the test, you are maxing out the CPU. You can’t run the test on the same router which you are testing. Find a 12 core Xeon PC to generate the Btest
Ran from a brand new Quad Core Computer (i7-2700k). Only used 5% of my CPU on the computer - 100% on the router.
Receive TCP max 550mbps
Send TCP max 120mbps
Receive UDP max 950mbps
Send UDP max 120mbps
Send/Receive TCP and UDP max 80mbps
Looks good except I thought the Full Duplex test would be much higher.
Why is there such a big difference of CPU use between TCP and UDP tests of the same speed?
I have an RB1200 set to Local/Remote Tx of 3M/3M and a UDP test uses 2% in profiler at most (overall CPU use is 4-9%). When I change it to TCP however overall CPU is always 100%.
One odd thing is that the top process when running a TCP test is not always btest. It seems to alternate between dhcp, ppp, hotspot, unclassified.
Normis is correct,you need to get two highend pc’s and connect them either side of the routers and use the mikrotik bandwidth test tool for windows and do the test through the routers (thats how client traffic would be passing through the routers anyway)
UDP is a connectionless protocol meaning it doesnt request the file again if parts or segments of it didnt reach the other side whereas TCP is a connection oriented protocol,the router acknowledges all parts of the file have been received,this in turn creates overhead traffic which explains the test results.For bandwidth tests always use UDP,that is the true test of any networking device.
You’re saying its unreasonable for me to expect that a RB1200 with a 999MHz CPU can generate and track 1 megabit of traffic without maxing out the CPU?
Just connect router to Internet with good uplink, connect your laptop and start downloading a torrent that have several thousands of seeds - you will get most realistic test possible.
In the lab test should consist of 3 devices - PC(source)<->router<->PC(destination)
i tried running btest on in between 2 pcs they never connect to each other while mik can connect to either of pcs made as btest server.
Digging up a decade old topic! congrats!
Try to set a limit to both receiving and transmitting. I have encountered when without a limit the program does not work. And note that TCP test version of the program with a file size of 750 kb is not supported. The old version of the program with a file size of 153kb does.
