----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/10268/#review18784 -----------------------------------------------------------
src/slave/slave.hpp <https://reviews.apache.org/r/10268/#comment39206> yay! src/slave/slave.cpp <https://reviews.apache.org/r/10268/#comment39203> Seems you can indeed use INITIALIZING, and then transition to RECOVERING here, no? src/slave/slave.cpp <https://reviews.apache.org/r/10268/#comment39204> Like like you should transition to TERMINATED after this wait() call. src/slave/slave.cpp <https://reviews.apache.org/r/10268/#comment39207> This looks strange to me, if we don't transition to REGISTERING, we shouldn't be registering.. A conditional guard on the transition implies the same conditional guard for the action. What was the motivation for the guard? Do we need to have registration state separate from the overall slave state? src/slave/slave.cpp <https://reviews.apache.org/r/10268/#comment39208> Ditto here on the guard. src/slave/slave.cpp <https://reviews.apache.org/r/10268/#comment39209> Should this be UNREGISTERED? It looks like we're not REGISTERING at this point. src/slave/slave.cpp <https://reviews.apache.org/r/10268/#comment39210> Looks like this CHECK will fail based on my above comments regarding the conditional transition. src/slave/slave.cpp <https://reviews.apache.org/r/10268/#comment39211> If we had a separate registration state, this could then be REGISTERED. Does that simplify things? src/slave/slave.cpp <https://reviews.apache.org/r/10268/#comment39212> Do we want a REREGISTERING state? src/slave/slave.cpp <https://reviews.apache.org/r/10268/#comment39213> Ok, this bit here makes me think we do need separate states for (re-)registration, since it's a somewhat independent process. - Ben Mahler On April 3, 2013, 6:26 p.m., Vinod Kone wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/10268/ > ----------------------------------------------------------- > > (Updated April 3, 2013, 6:26 p.m.) > > > Review request for mesos, Benjamin Hindman and Ben Mahler. > > > Description > ------- > > See summary. > > > Diffs > ----- > > src/slave/slave.hpp 2529bf500a3265b10ad4cddde10c2d62a6cdb4a0 > src/slave/slave.cpp 325231458a6883019436e7cc5a37f85f0f5735fa > > Diff: https://reviews.apache.org/r/10268/diff/ > > > Testing > ------- > > make check > > > Thanks, > > Vinod Kone > >
