What is the theoretical probability of an error in a packet of data
not being detected by a TCP checksum?   Please compare with the level of error
performance (ie where TCP checksum fails to detect error) that is typically
seen in a modern (high speed/ bandwidth) wide area networks? 
Please provide references to source material / data.

---

Clarification of Question by vp8csr-ga on 16 Nov 2004 09:58 PST

Thanks for the comment by lannaeus-ga.  Funilly enough, this was one
article that prompted my question. Unfortunately the data on which
this is based is a little old by internet time. To clarify then, the
data I am looking for should relate         more to the environment
typically found in a corporate wide area network and modern data
centre than less robust network infrastructures of a few years ago. 
Furthermore the final statement quoted should I think include the word
'fail'.  Even so, is billion 10E9 or 10E12 ? "After an analysis we
conclude that the checksum will [fail] to detect errors for roughly 1
in 16 million to 10 billion packets."

"Traces of Internet packets from the past two years show that between
1 packet in 1,100 and 1 packet in 32,000 fails the TCP checksum, even
on links where link-level CRCs should catch all but 1 in 4 billion
errors. For certain situations, the rate of checksum failures can be
even higher: in one hour-long test we observed a checksum failure of
1 packet in 400. We investigate why so many errors are observed, when
link-level CRCs should catch nearly all of them. We have collected
nearly 500,000 packets which failed the TCP or UDP or IP checksum.
This dataset shows the Internet has a wide variety of error sources
which can not be detected by link-level checks. We describe analysis
tools that have identified nearly 100 different error patterns.
Categorizing packet errors, we can infer likely causes which explain
roughly half the observed errors. The causes span the entire spectrum
of a network stack, from memory errors to bugs in TCP. After an
analysis we conclude that the checksum will to detect errors for
roughly 1 in 16 million to 10 billion packets."

Excerpt from 'When The CRC and TCP Checksum Disagree' - Stone, Partridge (2000)
http://citeseer.ist.psu.edu/update/433257

Hope it helps.

Lannaeus

Hi vp8csr,

What a coincidence! Yeah I couldn't find something more actual.

I think 1 billion should be here 10E9, because the author is from
Stanford University in USA.