Please refer to the installation instructions from Cassandra docs: http://cassandra.apache.org/doc/latest/cassandra/getting_started/installing.html

You may use either pipenv or manual installation to install the python dependencies.

a) Using pipenv

pip install pipenv
pipenv install
pipenv shell

b) Manual installation

pip install cassandra-driver

The data files are not included in this repository due to their large size. You may either download and extract them manually, or run the following commands in the project's root.

wget http://www.comp.nus.edu.sg/~cs4224/project_files_4.zip
unzip project_files_4.zip

Now, you should see a folder called project_files_4 in the project's root folder.

Cluster name: "CS4224B Cluster"	
listen_address: 192.168.48.169 (replace this with node ip)
rpc_address: 192.168.48.169 (replace this with node ip)
seeds: "192.168.48.169,192.168.48.170" 
native_transport_port: 6042
storage_port: 6000
ssl_storage_port: 6001

read_request_timeout_in_ms: 120000 
range_request_timeout_in_ms: 400000
write_request_timeout_in_ms: 50000
request_timeout_in_ms: 400000 

Set the default consistency level to LOCAL_QUORUM

self.consistency_level = cassandra.ConsistencyLevel.LOCAL_QUORUM

In create_db_tables.py, you can modify the variables for IP_ADDRESSES. For example, if you wish to use the IP addresses 192.168.48.169 and 192.168.48.170 on port 6042, you can change IP_ADDRESSES as shown below. If no port is supplied, the default port is 9042.

IP_ADDRESSES = [('192.168.48.169', 6042), ('192.168.48.170', 6042)]

If you wish to use a different consistency level than CONSISTENCY_LEVEL.ONE, please change the CONSISTENCY_LEVEL variable to the appropriate consistency level. For example:

CONSISTENCY_LEVEL = ConsistencyLevel.LOCAL_QUORUM

Then, ssh to each machine from xcnc20 to xcnc24

sshpass -p <cluster_password> ssh [email protected]

cassandra

cd temp/CS4224_Cassandra
python3 create_db_tables.py
python3 clean_data.py
cqlsh 192.168.48.169 6042 -f create_index.cql
cqlsh 192.168.48.169 6042 -f load_db_data.cql

bash run_experiment.sh <cluster_password> <experiment_type>

experiment_type should either be A or B

python3 consolidate_client_results.py
python3 compute_throughput_statistics.py
python3 compute_end_state_statistics.py

tail -f ~/temp/apache-cassandra-4.0.1/logs/system.log

nodetool status

tail -f nohup.out

ps ax | grep cassandra
kill <process id>