General technical discussions.
swinehold
New Member
Posts:
2
Joined:
Thu May 20, 2010 8:47 am
Certs:
None

iPerf/jPerf UDP vs TCP

Thu May 20, 2010 9:11 am

Hi,

We just upgraded one of our WAN links and we're trying to get an accurate measurement of the new throughput. We've been playing around with jPerf and we've been getting totally different results depending on whether we use TCP or UDP. For example, running a TCP test we get less than 1 Mbps throughput. Running a UDP test (with which we're able to set the amount of data we push) we're able to get throughput as high as 300 Mbps. I should point out that I was running this test to a server across the country with a round-trip latency of about 80 ms.

So which is the more accurate reflection of our throughput, TCP or UDP? I suspect the TCP is getting bogged down because of the high latency. Maybe TCP is sending packets up to its window size, then because of the latency it doesn't receive an ACK right away, so it's shrinking its window, and jPerf interprets this as window resizing due to a lack of available bandwidth.

UDP, on the other hand, will just pump out UDP packets and depending on the replies it gets from the server end, it will determine how many of those UDP packets were received. Based on that, it will determine bandwidth. So if that's the case, I would think UDP would be a purer reflection of your available bandwidth.

Any thoughts on this? Hopefully someone out there is an iPerf/jPerf expert (and also a TCP/UDP expert).

Thanks.

User avatar
Dinger
Post Whore
Posts:
1397
Joined:
Fri Apr 25, 2008 2:16 pm
Certs:
CCNP, CCNA:Sec, MCSE

Re: iPerf/jPerf UDP vs TCP

Thu May 20, 2010 10:08 am

I've never found the iperf UDP results to be accurate.

Definitely test with TCP; however, because of the high latency, you'll need to increase your TCP Window size in iperf (the -w option).

There are formulas out there that will tell you the ideal window size, given the bandwidth and delay (you listed the delay, but not the total bandwidth; I'm doubting 300 Megabyte/sec is how fast your link is!!).
"A problem well stated is a problem half solved". (Charles Kettering)

User avatar
mellowd
CCIE #38070
Posts:
13814
Joined:
Wed Jun 18, 2008 7:49 am
Certs:
CCIE (RS,SP), JNCIE-SP, BC-/SPNE/NP

Re: iPerf/jPerf UDP vs TCP

Thu May 20, 2010 10:36 am

We test using UDP

but we don't use iperf

swinehold
New Member
Posts:
2
Joined:
Thu May 20, 2010 8:47 am
Certs:
None

Re: iPerf/jPerf UDP vs TCP

Thu May 20, 2010 11:08 am

Thanks for the input. I ratcheted up the TCP window size and was able to get up to about 6 Mbps throughput, according to jPerf.

Any suggestions for other tools, free or otherwise? We run SolarWinds Network Performance Monitor, but as far as I've seen they don't offer a throughput tester.

User avatar
mellowd
CCIE #38070
Posts:
13814
Joined:
Wed Jun 18, 2008 7:49 am
Certs:
CCIE (RS,SP), JNCIE-SP, BC-/SPNE/NP

Re: iPerf/jPerf UDP vs TCP

Thu May 20, 2010 11:10 am


afouneh
New Member
Posts:
6
Joined:
Fri May 21, 2010 4:28 pm
Certs:
BSc in Electronics and Comms Eng.

Re: iPerf/jPerf UDP vs TCP

Sat May 22, 2010 5:38 pm

I have also used iPerf... and the results were very weird ... for UDP the bandwidth was 1 Mb/s and for TCP around 1 kb/s !
any ideas what may went wrong ?

User avatar
ibarrere
Cisco Inferno
Posts:
10283
Joined:
Mon Jul 10, 2006 12:58 am

Re: iPerf/jPerf UDP vs TCP

Mon May 24, 2010 7:43 pm

Dinger wrote:I'm doubting 300 Megabyte/sec is how fast your link is!!).


/byte/bit/

'

Return to General Tech

Who is online

Users browsing this forum: No registered users and 16 guests