Installing Apache Cassandra

  • https://www.digitalocean.com/community/tutorials/how-to-install-cassandra-and-run-a-single-node-cluster-on-a-ubuntu-vps

Cassandra (as well as Spark) requires Java to be installed. See the Spark installation documents to install Java.

Install & Setup Cassandra

Download the tarball and unzip it.

$ wget http://www.mirrorservice.org/sites/ftp.apache.org/cassandra/2.1.6/apache-cassandra-2.1.6-bin.tar.gz

$ tar -xvzf apache-cassandra-2.1.6-bin.tar.gz

Next, make sure that the folders Cassandra accesses, such as the log folder, exists and that Cassandra has the right to write on it:

$ sudo mkdir /var/lib/cassandra
$ sudo mkdir /var/log/cassandra
$ sudo chown -R $USER:$GROUP /var/lib/cassandra
$ sudo chown -R $USER:$GROUP /var/log/cassandra

In the .bash_profile add the following two lines:

$ nano ~/.bash_profile

# APACHE CASSANDRA Config.
export CASSANDRA_HOME=~/cassandra
export PATH=$PATH:$CASSANDRA_HOME/bin

Make sure to source ~/.bash_profile after this for the changes to take effect.

Running Cassandra on a single machine

To start Cassandra now we can use:

# Starting it as a standalone process
$ sudo sh ~/apache-cassandra-2.1.6/bin/cassandra

This will start it in the foreground. Lets quit this using CTRL+C. To make sure we don’t have it running in the foreground we can add two additional parameters nohup (disable all the stdout output) and & (push to background):

$ sudo nohup sh ~/apache-cassandra-2.1.6/bin/cassandra &
# Output looks something like:
[1] 1642
nohup: ignoring input and appending output to ‘nohup.out’

To stop it we first need to find out the process-id. We can do this using:

$ ps aux | grep cassandra
root      1704 36.5 29.8 1329036 1235776 pts/0 SLl  11:30   0:10 java -ea -javaagent:/home/fiordland/apache-cassandra-2.1.6/bin/../lib/jamm-0.3.0.jar -XX:+CMSClassUnloadingEnabled -XX:+UseThreadPriorities [...]

To kill this process use sudo kill -9 <PID> where the is 1704 in the example here.

While it is still running, lets connect locally using the CLI (command line interface):

# Starting the command-line-interface
$ sudo sh ~/apache-cassandra-2.1.6/bin/cassandra-cli
Connected to: "Test Cluster" on 127.0.0.1/9160
Welcome to Cassandra CLI version 2.1.6

The CLI is deprecated and will be removed in Cassandra 2.2.  Consider migrating to cqlsh.

Since this is depreciated we’ll move to the new interface. Exit the CLI using exit;. Now start the CQLSH using:

# Starting the cassandra-query-language shell
$ sudo sh ~/apache-cassandra-2.1.6/bin/cqlsh
Connected to Test Cluster at 127.0.0.1:9042.
[cqlsh 5.0.1 | Cassandra 2.1.6 | CQL spec 3.2.0 | Native protocol v3]

cqlsh> describe cluster;

Cluster: Test Cluster
Partitioner: Murmur3Partitioner

cqlsh> describe schema;

cqlsh> describe keyspaces;

system_traces  system

cqlsh> describe tables;

Keyspace system_traces
----------------------
events  sessions

Keyspace system
---------------
peers             schema_triggers   batchlog                 local
range_xfers       sstable_activity  size_estimates           hints
schema_keyspaces  peer_events       compaction_history
schema_columns    schema_usertypes  compactions_in_progress
"IndexInfo"       paxos             schema_columnfamilies

CQL - Cassandra Query Language?

  • http://www.planetcassandra.org/create-a-keyspace-and-table/
  • http://docs.datastax.com/en/cql/3.0/cql/cql_reference/insert_r.html
  • http://docs.datastax.com/en/cql/3.0/cql/cql_reference/cqlsh.html

So there is nothing there. Lets quickly check out how to create a keyspace and tables and insert some data before moving on the setting up other nodes and initialising a cluster.

Creating a keyspace with the name demo:

cqlsh> CREATE KEYSPACE demo WITH REPLICATION = { 'class' : 'SimpleStrategy', 'replication_factor' : 1 };

cqlsh> describe keyspaces;

system_traces  system  demo

We can now use this keyspace and create some tables there

cqlsh> use demo;
cqlsh:demo>

cqlsh> CREATE TABLE users (
firstname text,
lastname text,
age int,
email text,
city text,
PRIMARY KEY (lastname));

cqlsh:demo> describe schema

CREATE KEYSPACE demo WITH replication = {'class': 'SimpleStrategy', 'replication_factor': '1'}  AND durable_writes = true;

CREATE TABLE demo.users (
    lastname text PRIMARY KEY,
    age int,
    city text,
    email text,
    firstname text
) WITH bloom_filter_fp_chance = 0.01
    AND caching = '{"keys":"ALL", "rows_per_partition":"NONE"}'
    AND comment = ''
    AND compaction = {'min_threshold': '4', 'class': 'org.apache.cassandra.db.compaction.SizeTieredCompactionStrategy', 'max_threshold': '32'}
    AND compression = {'sstable_compression': 'org.apache.cassandra.io.compress.LZ4Compressor'}
    AND dclocal_read_repair_chance = 0.1
    AND default_time_to_live = 0
    AND gc_grace_seconds = 864000
    AND max_index_interval = 2048
    AND memtable_flush_period_in_ms = 0
    AND min_index_interval = 128
    AND read_repair_chance = 0.0
    AND speculative_retry = '99.0PERCENTILE';

Insert a record. Please not that for this table the lastname is the primary-key, so we cannot have more than 1 person with the same lastname.

cqlsh> INSERT INTO users (firstname, lastname, age, email, city) VALUES ('John', 'Smith', 46, 'johnsmith@email.com', 'Sacramento');

Setting up the Cluster

First thing we need to do is generate tokens for all machines in the cluster. Create a new python file to do this (following these instructions):

$ nano apache-cassandra-2.1.6/tokens.py

Paste in the following python code Note: Murmur3 codes.

#! /usr/bin/python
import sys
if (len(sys.argv) > 1):
    num=int(sys.argv[1])
else:
    num=int(raw_input("How many nodes are in your cluster? "))
# for i in range(0, num):
    # print 'token %d: %d' % (i, (i*(2**127)/num))
print [str(((2**64 / num) * i) - 2**63) for i in range(num)]

Run it to generate X-number of tokens.

$ python apache-cassandra-2.1.6/tokens.py

How many nodes are in your cluster? 2
token 0: 0
token 1: 4611686018427387904

Now we need to configure both machines so they recognize each other and there is one lead, or seed machine.

On both machines open the cassandra configuration file using

$ nano apache-cassandra-2.1.6/conf/cassandra.yaml

MACHINE 1 (IP: 172.16.0.223)

initial_token: 0

- seeds: 172.16.0.223 # ip address of 'seed node'

listen_address: 172.16.0.223

MACHINE 2 (IP: 172.16.0.221)

initial_token: 4611686018427387904

- seed: 172.16.0.223

listen_address: 172.16.0.221

Once you have saved both files. Start cassandra on both machines (maybe best to start machine 1 (seed-machine) first).

$ sudo nohup sh ~/apache-cassandra-2.1.6/bin/cassandra &

Check on both if the process is actually running and it did not error out somehow.

$ ps aux | grep cassandra

On Machine-1 now run the following command to see if it has picked up the other node in its network.

$ ./apache-cassandra-2.1.6/bin/nodetool status

Datacenter: datacenter1
=======================
Status=Up/Down
|/ State=Normal/Leaving/Joining/Moving
--  Address       Load       Tokens  Owns    Host ID                               Rack
UN  172.16.0.223  210.88 KB  256     ?       329edf4b-89b0-4d0d-bbbb-dd5f1962e936  rack1
UJ  172.16.0.221  46.81 KB   1       ?       6071a3bf-6dc3-4667-807e-925c2649d3e9  rack1

Note: Non-system keyspaces don't have the same replication settings, effective ownership information is meaningless

TO DO

To check if it is actually working… create a keyspace,table,insert on one and check on the other. (using the cqlsh cli)

Notes on keyspaces and replication/networktopology: http://docs.datastax.com/en/cassandra/1.2/cassandra/architecture/architectureDataDistributeReplication_c.html

CREATE KEYSPACE "demoShared" WITH REPLICATION = {'class' : 'NetworkTopologyStrategy', '172.16.0.223': 2, '172.16.0.221': 3};

NOT DONE

Other

From the DOCs

Is there a GUI admin tool for Cassandra?

  • DataStax Opscenter, a management and monitoring tool for Cassandra with a web-based UI.
  • Cassandra Cluster Admin, a PHP-based web UI.
  • Toad for Cloud Databases, a desktop application and Eclipse plugin which support Cassandra.
  • DBeaver, a desktop application, support Cassandra using JDBC driver.