See other failures under 000.html.
Three frame failures from one rider, all failures using 700C wheels, Rohloff Speedhubs, and disk brakes.
From http://forums.mtbr.com/showthread.php?t=325268, posted 2007/08 by "bryceb", available as of as of 2010/02.
[A]ll of the failures were of a similar nature. Two were on alloy FS 29'ers and one on a steel hardtail 29'er. I don't want to mention brands here as it is unfair to the manufacturers, but the hardtail was a custom and the builder was specifically asked to build to Rohloffs specs, but didn't and the bike eventually broke. He repaired the frame for me and followed Rohloffs specs, but a little to late!
The first of the FS alloy frames was sold to me on the belief of the supplier that the Rohloff would be fine, but wasn't, and the second (different manufacturer) meet a similar fate, after telling the builder of my woes and I had his absolute belief and guarrantee that there wouldn't be an issue. That frame broke after seven months, and he has subseqently re-modelled the chainstay. A sister bike bought at the same time broke in exactly the same place after two years of riding, but he doesn't do as much as I, or is as hard on gear apparantly.
We ride a lot of very steep stuff and it seems that the braking forces together with the Rohloff torque input, and quite possibly the extra stresses impossed by the larger wheel amplifying the torque enhances the problem.
It is unclear what role, if any, the Rohloff played in these failures. Braking forces should typically be much larger than Rohloff torque arm reaction forces, but braking is relatively infrequent and likely fairly smooth, while torque arm reaction forces repeat and vary with every pedal stroke, thus may give rise to fatigue where larger braking forces usually do not.
See also FAIL-138.html.