Installing a Cassandra Cluster Step By Step

⚙️ Prerequisites

• Operating System: Cassandra works well on Linux distributions like Ubuntu, CentOS, or Red Hat.
• Java: Cassandra requires Java 8 or newer. Ensure Java is installed on each server in the cluster.
• Servers: At least two or more servers are needed for a cluster setup (to demonstrate replication and distribution).
• Networking: All servers should be networked with each other. Ensure that required ports (7000, 9042, and others) are open.

🔧 Step 1: Setting Up the Environment

Update Packages on all nodes:

sudo apt-get update
sudo apt-get upgrade

Install Java (if not already installed):

sudo apt-get install openjdk-11-jdk -y

Confirm Java installation:

java -version

The output should show a version of Java 8 or newer.

📥 Step 2: Download and Install Apache Cassandra

Add the Cassandra Repository:

echo "deb https://debian.cassandra.apache.org 311x main" | sudo tee -a /etc/apt/sources.list.d/cassandra.sources.list
wget -q -O - https://debian.cassandra.apache.org/debian/repo_key | sudo apt-key add -

Install Cassandra:

sudo apt-get update
sudo apt-get install cassandra -y

Verify Installation:

sudo systemctl status cassandra

Check the running Cassandra process:

nodetool status

🛠️ Step 3: Configure Cassandra for Cluster Setup

Edit the primary configuration file cassandra.yaml located in /etc/cassandra/:

sudo nano /etc/cassandra/cassandra.yaml

Modify the following parameters:

• • Cluster Name: Set the same cluster name on all nodes.

cluster_name: 'MyCassandraCluster'

• • Seeds: Specify the IP addresses of seed nodes.

seed_provider:
  - class_name: org.apache.cassandra.locator.SimpleSeedProvider
    parameters:
      - seeds: "192.168.1.1,192.168.1.2"

• • Listen Address: Set to the node’s IP address.

listen_address: 192.168.1.1

• • RPC Address: Set to the node’s IP address or 0.0.0.0:

rpc_address: 0.0.0.0

• • Endpoint Snitch: For a simple cluster setup, use:

endpoint_snitch: SimpleSnitch

Save and close the file. Restart Cassandra to apply changes:

sudo systemctl restart cassandra

✅ Step 4: Verify the Cluster

Use nodetool status to check the cluster’s status:

nodetool status

The output should show the cluster’s nodes and their statuses.

📂 Step 5: Create a Keyspace and Test Replication

Open the Cassandra Command Line Interface (cqlsh):

cqlsh 192.168.1.1

Create a Keyspace:

CREATE KEYSPACE test_keyspace WITH REPLICATION = {'class': 'SimpleStrategy', 'replication_factor': 2};

Use the Keyspace and Create a Table:

USE test_keyspace;
CREATE TABLE users (user_id UUID PRIMARY KEY, name TEXT, email TEXT);

Insert Data and Query:

INSERT INTO users (user_id, name, email) VALUES (uuid(), 'John Doe', '[email protected]');
SELECT * FROM users;

📡 Step 6: Troubleshooting and Seed Connection

Seed Connection Explanation:

Seed nodes act as the entry point for new nodes joining the cluster. When a node starts, it contacts the seed nodes to learn about the cluster topology. Ensure the following for proper connectivity:

• All nodes can ping the seed nodes by their IP addresses.
• Ports 7000 (intra-node communication) and 9042 (client communication) are open on all nodes.
• The seed nodes have consistent cassandra.yaml configurations with accurate seed IPs.

Troubleshooting Steps:

It seems the troubleshooting section is still lacking substantial additions to meet your expectations. Let me create a fully detailed troubleshooting guide and ensure it enhances the existing document appropriately.

Detailed Troubleshooting Additions:

1. Seed Node Connectivity Issues:
   • Ping Test: Confirm that all nodes can communicate with the seed nodes:

     ping <seed-node-IP>
     

     If the ping fails, verify network configurations and firewalls.

   • Ports Test: Ensure ports 7000 (intra-node communication) and 9042 (client communication) are open. Use the nc command:

     nc -zv <seed-node-IP> 7000
     

   • DNS Resolution: If using hostnames instead of IPs for seeds, ensure the DNS resolves correctly:

     nslookup <seed-node-hostname>
     

2. Common Errors and Fixes:
   • Error: Node Not Joining Cluster
     • Cause: Incorrect seed IPs or mismatched cluster name.
     • Fix: Ensure the same cluster_name is set in all cassandra.yaml files, and the seeds list is accurate.
   • Error: JVM Heap Size Issues
     • Cause: Memory exhaustion.
     • Fix: Modify the heap size settings in /etc/cassandra/jvm.options:

       -Xms4G
       -Xmx4G
       

   • Error: Read/Write Timeout
     • Cause: High load or network latency.
     • Fix: Increase timeouts in cassandra.yaml:

       read_request_timeout_in_ms: 10000
       write_request_timeout_in_ms: 10000
       

3. Logs and Debugging:
   • Use Cassandra logs for detailed error messages:

     sudo tail -f /var/log/cassandra/system.log
     

     Look for messages about gossip, seed discovery, or schema disagreements.

4. Testing Cluster State:
   • Validate the cluster setup with:

     nodetool status
     

     Check the UN (Up and Normal) status for all nodes. If nodes are down, revisit configurations.