Setting up environment

Required: JRuby 9.1.5.0 (Ruby 2.3.1), Rails 4.2.1

Install JRuby (using RVM)

1. If no RVM is installed, install it with:

\curl -sSL https://get.rvm.io | bash -s stable

More information: https://rvm.io/rvm/install

2. Install specific JRuby version. Go to app root directory and run in console:

rvm use jruby-9.1.5.0 --install

Install all the gems

bundle install

Starting up the splicemachine and the app

First, make sure splice is up and running:

1. cd ~/spliceengine/

2. ./start-splice-cluster (only for the first time). Run ./start-splice-cluster -bon every other time

`./start-splice-cluster -h`, for any addtional information

Possible issues with missing libary:

If there are issues with running a splice cluster, you might need to add next line into your ~/.bashrc file:

export LD_LIBRARY_PATH=/usr/local/lib

Possible issue with freezing migrations:

Splice requires a certain amount of free space on the hard-drive in order to be run correctly. Make sure that there is at least 10% of free space on the hard drive. If you have for example 80 GB, make sure that at least 8 GB is available of free space, as 2-3 GB might not be enough. (The actual free space might not be used, but the hard-drive will need to have it)

Notes:

On linux, if splice cluster is not run on the first time (after running a ./start-splice-cluster -b), stop the execution, and run ./start-splice-cluster -b again.

Setting up a database

Now when the cluster is up and running, we need to set-up the database:

1. Delete the config/database.yml file.
2. Locate database_example.yml and rename it to database.yml
3. Run bundle exec rake db:migrate

Set-up production environment (needed for testing puma in production mode)

1. Open a consolsole and run bundle exec rake secret and copy the generated string
2. Open .bashrc and write at the end: export SECRET_KEY_BASE=COPY_AND_PASTE_STRING_GENERATED_FROM_STEP_1
3. Restart computer, or just for this session, run in console source ~/.bashrc

Note: In an error about uglifier appears, you might need to install a nodejs on linux:

$ sudo apt-get install nodejs

Additional information (not neeeded for the set-up)

Gem which is needed in order to use splice is:

gem 'activerecord-jdbcsplice-adapter'

And, it depends on:

gem 'jdbc-splice' (no need to specify it directly, it is done in the background)

Testing

Before running tests or benchmarks, make sure that splice cluster is running: ./start-splice-cluster -b

Run all test tests

bundle exec rspec spec/

Run specific test

bundle exec rspec spec/models/company_spec.rb:13 (13 is the line number)

Run specific test with logs enabled

LOGS=true bundle exec rspec spec/models/company_spec.rb:13

Run real server-client test:

Note: Some of the SQL commands will assume that there are records in database, like where, update etc. , so in bundle exec rails c, and create a single record with: Company.create(name: 'Company') before running the tests.

Please note the ID of the created record, as it will be used in some of the tests.

1. Run the server: bundle exec puma -p 3000 -t 16:16 -e production

   Note: Run server with logs enabled: LOGS=true bundle exec puma -p 3000 -t 16:16 -e production

2. Run the benchmark with any of these commands:

   • ab -n 10000 -c 1000 -r http://localhost:3000/benchmarks/method_where?id=NOTED_ID (special case)
   • ab -n 10000 -c 1000 -r http://localhost:3000/benchmarks/method_create
   • ab -n 10000 -c 1000 -r http://localhost:3000/benchmarks/method_update
   • etc.

Note: On mac, there is an issue with Apache Benchmark. In order to fix it, follow instructions from http://sudheeraedama.blogspot.rs/2013/05/install-apachebench-on-mac-osx-1075.html

Methods supported:

method_create, method_update, method_where, method_limit, method_offset, method_group, method_select

Run benchmarks on single SQL commands (where, create, limit etc.):

bundle exec rake benchmark:models RAILS_ENV=test

Switching between Ruby and JRuby

In this app Ruby is used for MySQL and JRuby for Splice

If you want to run the code against ruby code and test the app with it, open file .ruby-version and change any text in it to ruby-2.3.1, or any other ruby version. Return to the previous directory with (cd ..), and re-enter the app directory by cd splice-jruby (in order to refresh the settings)

The database settings for ruby is in database_ruby.yml. No need to replace or delete the database.yml, just modify the database_ruby.yml file with your settings.

Helper Methods

In order to drop all tables in current database, run in console:

bundle exec rake db:splice:destroy_all

Implementation of Rails on CentOS (centos-release-6-8.el6.centos.12.3.x86_64)

• sudo yum groupinstall -y 'development tools'
• gpg2 --keyserver hkp://keys.gnupg.net --recv-keys 409B6B1796C275462A1703113804BB82D39DC0E3
• curl -L get.rvm.io | bash -s stable (In case the keys are not present then the above key would be shown )
• source ~/.profile
• rvm install jruby in case ruby gems is not installed (rvm ruby gems latest)
• gem install rails -v 4.0.0
• gem install bundle
• git clone https://github.com/splicemachine/splice-community-sample-code.git
• navigate to “tutorial-rails” and run "bundle install"
• which jruby whatever directory comes substitute bin for lib and put the client driver file there. In my case i for ~/.rvm/rubies/jruby-9.0.5.0/bin/jruby ~/.rvm/rubies/jruby-9.0.5.0/lib