redis cluster configuration

redis_cluster_config.sh

# 212.71.252.54  / 192.168.171.141 / node1
# 176.58.103.254 / 192.168.171.142 / node2
# 178.79.153.89  / 192.168.173.227 / node3

# local hosts
192.168.171.141 node1
192.168.171.142 node2
192.168.173.227 node3

# remote hosts
212.71.252.54  node1
176.58.103.254 node2
178.79.153.89  node3

mkdir build
cd build
wget http://download.redis.io/releases/redis-3.0.0.tar.gz
tar -xvzf redis-3.0.0.tar.gz
cd redis-3.0.0/

apt-get install -y make gcc build-essential
make MALLOC=libc # TODO: test with jemalloc

# not neccessary, actually
apt-get install -y tk8.5 tcl8.5
make test

# make 'bind' on redisN
# enable cluster in config
src/redis-server ./redis.conf


# ruby on bootstrap server (node1)
add-apt-repository ppa:brightbox/ruby-ng-experimental
apt-get update
apt-get install -y ruby2.2

# run on bootstrap server, ips instead of local names (at least for now)
src/redis-trib.rb create 192.168.171.141:6379 192.168.171.142:6379 192.168.173.227:6379
# if something is wrong
src/redis-trib.rb check redis1:6379
src/redis-trib.rb fix redis2:6379

# just for testing
wgwt https://github.com/antirez/redis-rb-cluster/archive/master.zip
unzip master.zip

cd redis-rb-cluster
ruby example.rb # k-c

ruby consistency-test.rb # and leave
# on node2
src/redis-cli -c -h node2 debug segfault
# cluster is down, good job!