Replication for fault-tolerance: what have we learned in forty years?

André Schiper presenting his «Leçon d'honneur»

Last Friday I went back to EPFL to listen to the last lecture of my PhD advisor, André Schiper. His Leçon d’honneur had the title Replication for fault-tolerance: what have we learned in forty years?. The school needed to change the room of the presentation multiple time as more and more people registered for the talk. In the end we ended up in one of the large amphitheatres in the Centre Ouest. This was a very moving event, many of André’s PhD students and Postdocs came for the occasion.

The presentation was a rare moment: it is quite rare in computer science to look at an issue from a 40 years perspective, from the pioneering papers to the maturity and use in the industry. This was also a reminder that research is not a straight line: there has been a lot of confusion, abstractions and models which ended up being irrelevant or useless, important papers that were ignored or misunderstood, problematic ones that ended up stealing the spotlight. This lecture certainly gave me a sense of closure on what I did during my research career and I hope the slides will be available online soon.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.