----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/62893/#review187674 -----------------------------------------------------------
src/slave/flags.cpp Lines 454-457 (patched) <https://reviews.apache.org/r/62893/#comment264727> Agree that it should be close, but disagree about the waste. The timer for `agent_reregister_timeout` only running when the master is running. In a situation when the master is down (due to a severe outage for example), the agent reregister timer is not running, hence it might make sense to have `recovery_timeout` longer than `agent_reregister_timeout`. - Alexander Rukletsov On Oct. 11, 2017, 2:11 p.m., Benno Evers wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/62893/ > ----------------------------------------------------------- > > (Updated Oct. 11, 2017, 2:11 p.m.) > > > Review request for mesos and Alexander Rukletsov. > > > Repository: mesos > > > Description > ------- > > Expand description of the --recovery_timeout flag. > > > Diffs > ----- > > src/slave/flags.cpp d424ade99bccde0c4c4fb591c8a443ccb8fd58ed > > > Diff: https://reviews.apache.org/r/62893/diff/1/ > > > Testing > ------- > > > Thanks, > > Benno Evers > >
