Boost logo

Boost :

Subject: Re: [boost] [asio] Randomly having problems reading from the socket
From: FlorianGeorge (fgeorge_at_[hidden])
Date: 2012-04-02 17:11:04


I am also having the problem that in my application, sending/receiving at
20Hz over a localhost/loopback connection, sometimes stalls for exactly
5000ms. This happens mostly on slower (e.g. Athlon X2 4400+) than on faster
(e.g. i7-930) machines.

Torben Wiggerich wrote
>
>> Have you put an analyzer (e.g. WireShark) on it and watched the TCP
>> packets from right before and when the pause occurs? This might help
>> isolate the problem.
>
> Ok I will try this...
>
Did this help finding the source of the problem?

As explained here (http://wiki.wireshark.org/CaptureSetup/Loopback),
Wireshark/WinPcap will not be able to capture traffic over the loopback
device. As suggested, I used RawCap (http://www.netresec.com/?page=RawCap)
instead.

I filter the data using "tcp.port == 1234" to not get sidetracked by other
traffic.

I uploaded a screenshot here: http://i.imgur.com/36U62.jpg
I uploaded the whole dumpfile here:
http://www8.zippyshare.com/v/50356468/file.html

The interesting part is probably this:

> 85030 16.690429 127.0.0.1 127.0.0.1 TCP 1500 [TCP segment of a reassembled
> PDU]
> 85041 21.689453 127.0.0.1 127.0.0.1 TCP 808 [TCP Previous segment lost]
> [TCP segment of a reassembled PDU]
>

You can clearly see the ~5000ms pause between those two events.

Does anyone have further insight in the reason why it is exactly 5000ms, and
how to maybe fix or work around it?

--
View this message in context: http://boost.2283326.n4.nabble.com/asio-Randomly-having-problems-reading-from-the-socket-tp3490016p4527192.html
Sent from the Boost - Dev mailing list archive at Nabble.com.

Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk