iDunno

Join IDunno

At machine 3 to 10, under the repo folder, execute command

./commands.sh join

to join IDunno. You will need superuser privilege to for this.

Only when at VM1, execute command

python3 Coordinator.py &
python3 CoordinatorML.py &

Since VM1 is the default Coordinator.

Only when at VM2, execute command

python3 StandByCoordinator.py &

Since VM2 is the stand-by Coordinator.

Submit job to IDunno

At any machine that have joined IDunno, execute command

python3 tools.py IDunnoSUB job_type sdfsfilename batch_size

Where job_type is either 'V' or 'T', standing for 'vision' and 'tranlation', sdfsfilename is the data to be inferenced batch_size is the batch size.

Start training

At any machine that have joined IDunno, execute command

python3 tools.py IDunnoMSG STR

Start inferencing

At any machine that have joined IDunno, execute command Do Not Expect too much, this is running on 10 VMs that was deployed in 2015, WTF do you expect???

python3 tools.py IDunnoMSG SINF

End jobs and collect result

At any machine that have joined IDunno, execute command

python3 tools.py IDunnoMSG EJ

Commands

At any machine that have joined IDunno, for commands C1, C2, C4, C5, execute command,

./commands.sh Cn

Where n is in {1,2,4,5}.

For C3, execute command,

./commands.sh C3 JobID batch_size

Where JobID the ID of Job you want to set batch_size of, batch_size is the batch size. Indy Gupta to dziwka