You need to create the --work_dir first (in your case '/tmp/work_dir') since the Mesos master doesn't create it automatically. Sorry about the error message, it looks much worse than it is!
This has since been fixed and will be included in 0.19.1. Ben. On Fri, Jun 27, 2014 at 1:47 AM, Diptanu Choudhury <[email protected]> wrote: > Hi, > > When I start Mesos 0.19.0 on OSX 10.9, I get the following error - > > ./bin/mesos-master.sh --zk=zk://localhost:2181/mesos > --work_dir=/tmp/work_dir/ --quorum=2 > > I0627 01:45:09.698272 1979613968 main.cpp:150] Build: 2014-06-27 01:03:13 > by dchoudhury > > I0627 01:45:09.698837 1979613968 main.cpp:152] Version: 0.19.0 > > F0627 01:45:09.701522 1979613968 replica.cpp:721] CHECK_SOME(state): IO > error: ~/work_dir/replicated_log/LOCK: No such file or directoryFailed to > recover the log > > *** Check failure stack trace: *** > > @ 0x10cb0cdc4 google::LogMessage::SendToLog() > > @ 0x10cb0d2f3 google::LogMessage::Flush() > > @ 0x10cb1048f google::LogMessageFatal::~LogMessageFatal() > > @ 0x10cb0d9e9 google::LogMessageFatal::~LogMessageFatal() > > @ 0x10c535559 _CheckFatal::~_CheckFatal() > > @ 0x10c535429 _CheckFatal::~_CheckFatal() > > @ 0x10c9c3dba mesos::internal::log::ReplicaProcess::restore() > > @ 0x10c9c3861 > mesos::internal::log::ReplicaProcess::ReplicaProcess() > > @ 0x10c9c6f9b mesos::internal::log::Replica::Replica() > > @ 0x10c93730a mesos::internal::log::LogProcess::LogProcess() > > @ 0x10c93cc80 mesos::internal::log::Log::Log() > > @ 0x10c93cbfd mesos::internal::log::Log::Log() > > @ 0x10c45180c main > > @ 0x7fff8b0465fd start > > Abort trap: 6 > > > Any ideas, with what's going on here? I hope I am not invoking > mesos-master incorrectly. > > -- > Thanks, > Diptanu Choudhury > Web - www.linkedin.com/in/diptanu > Twitter - @diptanu <http://twitter.com/diptanu> >
