NOTE!!!.txt

THIS ENVIRONMENT IS TESTED! 

Last edit:
Feb 9, 2018, Liang


For Shared-Memory, just click run button in IDE directly. The web-google dataset is used for pagerank algorithm test. One thing should noticed that every node should have out-degree, otherwise the incorrect result or crash may occurred.

For distributed test, 
open the terminal,
1. cd socialite
2. bin/run.py examples/PageRank/Google.dl

WARN:
1. YOU MUST CONFIGURE THE VMWare EQUALS OR MORE THAN 2 CPU CORES, ORTHERWISE THE SOCIALITE WILL REPORT EXCEPTION. THIS IS SOCIALITE BUG BUT NOT CAUSED BY MYSELF.
2. THIS ENVIRONMENT JUST FOR TEST, DO NOT PROCEED ANY PERFORMANCE EXPERIMENTS ON THE VM MACHINE.

INSTRUCTIONS

Configure distributed environment:

1.compile openmpi 2.1.2
  download and extract openmpi-2.1.2.tar.gz
  ./configure --prefix /opt/openmpi-2.1.2 --enable-mpi-java
  make all
  sudo make install

2.install JDK
  sudo add-apt-repository ppa:webupd8team/java
  sudo apt-get update
  sudo apt-get install oracle-java8-installer

3.configure environment variables
  vim ~/.bashrc
  append to the '.bashrc' file:

  export JAVA_HOME=/usr/lib/jvm/java-8-oracle
  export SOCIALITE_PREFIX=/home/neu/socialite
  export MPI_HOME=/opt/openmpi-2.1.2
  export PATH=$PATH:$JAVA_HOME/bin:$MPI_HOME/bin

4.edit machines file
  machines file tells MPI where and how many process to start-up. The first line of conf/machines describe the master hostname and process number to run. e.g. "ubuntu slots=5" means start 5 process on the machine "ubuntu", including 1 master process and 4 worker processes.
  Another example:
########################
  master slots=9
  hadoop0 slots=4
  hadoop1 slots=4
  hadoop2 slots=4
  hadoop3 slots=4
########################
  This file means there are total 25 processes to run, including 1 master process run on master, and 24 worker process run across from master to hadoop3

5.about dataset
  Every node should have out-degree (mentioned above)
  For distributed experiments, you can put the datasets on every machine in the same path or just put the datasets on hdfs.

6.about debugging and formal experiment
  For debugging, you should edit bin/common.py line 16, to tell the java to run the bytecodes compiled by Intellij IDEA. AFTER EDITED THE CODE, YOU SHOULD CLICK Build-Build Project TO GENERATE LATEST BYTECODE!!!! AND THEN RUN THE bin/run.py SCRIPT.
  For formal experiment:
  1. cd socialite
  2. ant compile
  3. edit the bin/common.py line 16, replace with "ENTRY_CLASS_PATH = SOCIALITE_PREFIX + '/classes/socialite.jar'"