rohc team mailing list archive

[Didier Barvaux <question181392@xxxxxxxxxxxxxxxxxxxxx>]

Question #181392 on rohc changed:
https://answers.launchpad.net/rohc/+question/181392

    Status: Open => Answered

Didier Barvaux proposed the following answer:
Hello,

I reformulate your problem to be sure to correctly understand it:
 - you send several UDP streams to machine A across a ROHC
   compressor/decompressor pair,
 - a regular interval of time the decompressor reports a CRC failure
   while decompressing a ROHC packet.

Is it correct?

To debug this kind of problem, you may:
 - Capture the uncompressed UDP streams before they go across the
   compressor/decompressor pair. You may use tcpdump for example
   (with tcpdump please use option -s0 to avoid to truncate packets).
 - Start the UDP streams that cause the problem.
 - Wait for the problem to be reproduced.
 - Stop the decompressor and the network capture.
 - Look at the decompressor traces. Search for the CRC failure. Collect
   information about the paquet in the traces (IP addresses, UDP ports...).
   You should also find a dump of the (malformed) decompressed packet.
 - Try to identify the corresponding packet in the network capture.
 - When found, compare the packet from the capture with the one dumped
   in the decompressor traces.

The differences you may find between the 2 packets often give some information
about the problem. Contact me on the mailing list if you need some help for
further analysis.

Regards,
Didier

-- 
You received this question notification because you are a member of ROHC
Team, which is an answer contact for rohc.