 For my <a href="http://www.cs.cornell.edu/Info/People/chandra/Thesis.html">Ph.D. thesis</a>, I studied unreliable failure detection in asynchronous distributed systems (together with <a href="http://www.cdf.toronto.edu:/DCS/CSRI/Hadzilacos.html">Vassos Hadzilacos</a> and my advisor <a href="http://www.cs.cornell.edu/Info/People/sam/sam.html">Sam Toueg</a>).  We showed that a surprisingly weak failure detector denoted <IMG  ALIGN=BOTTOM ALT="" SRC="http://www.cs.cornell.edu/Info/People/chandra/icons/diamondW.gif"> is sufficient to solve Consensus and Atomic Broadcast, two fundamental problems that cannot be solved in asynchronous distributed systems (click <a href="http://www.cs.cornell.edu/Info/People/chandra/UnreliableFD.html">here</a> to find out more). We then showed that <IMG  ALIGN=BOTTOM ALT="" SRC="http://www.cs.cornell.edu/Info/People/chandra/icons/diamondW.gif"> is the weakest failure detector for solving these problems in asynchronous systems (click <a href="http://www.cs.cornell.edu/Info/People/chandra/WeakestFD.html">here</a> to find out more).  For a summary of research on failure detection, click <a href="http://www.cs.cornell.edu/Info/People/chandra/FailureDetectionPapers.html">here</a>.<p>
