Demonstration looked very sharp, will need to pay attention to how it performs in production environments.

It is in production, and performing very well for us at the Biodiversity Heritage Library. I’ll save web stats for someone else to quote, but my implementation has run for weeks at a time without any intervention. The few issues we’ve come across have been addressed by Ryan (via the djatoka mailing list), and we’re happy being a test case to not only showcase the functionality of djatoka, but also serving as a testcase to drive this open source project to be more functional, stable and usable for others.

My initial implementation here:
http://www.fak3r.com/2009/01/27/howto-serve-jpeg2000-images-with-a-scalable-infrastructure/

I will be writing a followup detailing scaling and failover methods I going to put in place in a weeks time.

Thanks

P