Hi,

TL;DR: Upgrade to linux-3.13 to resolve the problem. It is not a tinc related 
problem.

A similar behavior was reported by James Page back in october 2013 
http://www.gossamer-threads.com/lists/openstack/dev/31628?do=post_view_flat. 
Although I was not able to find a definitive answer, it hinted that the problem 
was related to a kernel bug. I was running a 3.11 kernel with OpenVSwitch 
1.10.2. Upgrading to linux-3.13 with the in-tree version of OpenVSwitch 
resolved the problem. I switched back to 3.11 and OpenVSwitch 1.10.2 to check 
that the problems shows again and it did. The kernel upgrade has been verified 
to be the only modification. 

Cheers

On 29/04/2014 17:33, Loic Dachary wrote:
> It turns out that uploading with rsync works as fast as expected. The problem 
> therefore seems to be : uploading works at 100Mb/s and downloading is 
> degraded to 1Mb/s. Note that I checked that downloading on the same host 
> without involving GRE or tinc indeed delivers 100Mb/s.
> 
> This is puzzling :-)
> 
> On 29/04/2014 16:31, Loic Dachary wrote:
>> Hi,
>>
>> In a setup where OpenVSwitch is used with GRE tunels on top of an interface 
>> provided by tinc, I'm experiencing significant performance degradation 
>> problems (from 100Mb/s down to 1Mb/s in the worst case) and I'm not sure how 
>> to fix this. The manifestation of the problem is, from the user point of 
>> view, iperf reports ~100Mb/s and rsync reports ~1Mb/s:
>>
>> $ iperf -c 91.224.149.132
>> ------------------------------------------------------------
>> Client connecting to 91.224.149.132, TCP port 5001
>> TCP window size: 23.5 KByte (default)
>> ------------------------------------------------------------
>> [  3] local 10.0.2.27 port 57143 connected with 91.224.149.132 port 5001
>> [ ID] Interval       Transfer     Bandwidth
>> [  3]  0.0-10.0 sec  99.4 MBytes  83.3 Mbits/sec
>>
>>
>> $ rsync --progress 
>> [email protected]:precise-server-cloudimg-amd64-disk1.img .
>> precise-server-cloudimg-amd64-disk1.img
>>     32112640  12%   83.44kB/s    0:45:34
>>
>> A tcpdump of iperf ( http://dachary.org/loic/good.cap ) shows no 
>> retransmission while a tcpdump of rsync  ( http://dachary.org/loic/bad.cap ) 
>> shows a frequent "TCP Dup ACK", "TCP Retransmission", "TCP Out-Of-Order". 
>>
>> I'm not sure why this happens, any hint would be greatly appreciated :-)
>>
>> Cheers
>>
> 

-- 
Loïc Dachary, Artisan Logiciel Libre

Attachment: signature.asc
Description: OpenPGP digital signature

_______________________________________________
tinc mailing list
[email protected]
http://www.tinc-vpn.org/cgi-bin/mailman/listinfo/tinc

Reply via email to