[Interest] TCP ACK with QDataStream -- OR: disconnect race condition detection

Constantin Makshin cmakshin at gmail.com
Tue Sep 11 19:17:20 CEST 2012


Sorry for sending that message directly to you, Thiago — forgot to fix
the address before hitting the "Send" button. :)

-------- Original Message --------
On terça-feira, 11 de setembro de 2012, às 16.33.44, you wrote:
> There's a paper
> (http://conferences.sigcomm.org/sigcomm/2000/conf/paper/sigcomm2000-9-1.pdf)
> where estimations show that "between one (data) packet in 16 million
> packets and one packets in 10 billion packets will have an undetected
> checksum error". And th number of 10^10 packets doesn't look exceptionally
> high considering modern amount of [global] network traffic.
>
> That's what makes me think that in cases where data integrity is critical
> one shouldn't rely on TCP checksum alone and using additional
> application-level validations is not a bad idea.

True. That doesn't mean the TCP checksum is inappropriate for what it's
supposed to do. The aggregate protection of Ethernet/ATM, IP and TCP's
checksums and error correcting / retransmission measures will give you
probably a protection against a BER of 1e-9.

If your use-case requires better protection than that, you need to apply
it on
top.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 490 bytes
Desc: OpenPGP digital signature
URL: <http://lists.qt-project.org/pipermail/interest/attachments/20120911/1e8be25d/attachment.sig>


More information about the Interest mailing list