I am preparing myself for next week's High Performance Transaction Systems workshop in Asilomar. As you can see from the agenda I am presenting in the session on Continuous Processing Architectures. I was a bit ignorant when I agreed to present in this session as I didn't know that this was a term coined by the Standish Group (Jim Johnson is leading this session), to describe what we used to call active replication systems. Which I thought the database community had declared evil territory ever since Jim Gray's 1997 paper describing it as almost impossible to achieve when scale was involved.
There is little or no information out on CPA beyond the Standish white papers. I would be very happy to receive some pointers to specific criticisms or words of support for their architectural vision. Of course I can do a presentation on how I think one should build these systems using our wonderful Cornell technologies, but I would like a bit more balanced view. Any hints?
Posted by Werner Vogels at October 9, 2003 08:32 AM