Managing node count in Jira Data Center

The number of nodes guardrail measures the size and configuration of your cluster. An improper node count can affect load distribution, system performance, and redundancy.

Impact of node count on system health

The number of nodes in your Jira Data Center cluster can affect your system in several ways:

1. Redundancy and availability: having at least two nodes ensures high availability and enables zero downtime upgrades.

2. Performance and scalability: adding nodes can improve performance up to a point, after which inter-node communication overhead may degrade performance.

3. Resource utilization: proper node count helps optimize CPU, memory, and network resource usage.

4. Maintenance flexibility: additional nodes allow for easier maintenance and upgrades without system downtime.

What's the recommendation?

To ensure optimal performance, you should aim to configure your node count within the recommended limits. Here are the steps to maintain an effective node configuration:

• Ensure minimum redundancy

• Optimize node count for performance

• Configure specialized nodes

Ensure minimum redundancy

If you have fewer than two nodes, add at least one more node to your cluster for redundancy and zero downtime upgrades.

To add a node:

1. Stop one existing node

2. Copy the installation directory and local home directory to the new node

3. Edit the cluster.properties file in the local home directory:

   • Change jira.node.id to a new, unique identifier

4. Start Jira on the new node

5. Verify the new node:

   • Go to Administration > System > Clustering

   • Check that the new node is visible

Learn more about setting up a Jira Data Center cluster

For detailed steps on adding and removing nodes, see adding and removing Data Center nodes

Learn more about setting up a Jira Data Center cluster

For detailed steps on adding and removing nodes, see adding and removing Data Center nodes

Optimize node count for performance

If you have more than four nodes, it's important to evaluate your performance metrics to determine if reducing node count could improve overall system performance.

Key metrics to monitor:

1. CPU utilization across nodes

2. Response times for key operations (e.g., issue creation, search)

3. Network traffic between nodes

4. Cache replication times

5. Database connection pool utilization

Steps to optimize:

1. Collect performance data over a significant period (e.g., 2-4 weeks).

2. Analyze the data to identify bottlenecks or underutilized resources.

3. If inter-node communication appears to be a bottleneck, consider reducing node count.

4. Implement changes gradually, monitoring performance closely after each adjustment.

Consider vertical scaling (increasing resources per node) instead of adding more nodes if you're experiencing performance issues.

Find more about adding and removing Data Center nodes.

Configure specialized nodes

Dedicating specific nodes for certain tasks can optimize performance in large or complex Jira instances.

Recommended specialized node configuration:

1. Administration node: Dedicate one node to administrative tasks and maintenance operations. Exclude this node from the load balancer to prevent user traffic.

2. API nodes: if your instance experiences heavy API usage, consider dedicating 1-2 nodes specifically for API traffic.

3. Configure load balancer: Adjust load balancer settings to distribute traffic appropriately.

Steps to implement:

1. Identify the nodes you want to specialize.

2. Adjust your load balancer configuration to direct appropriate traffic to these nodes.

3. Monitor performance to ensure the specialization is improving overall system health.

Read more about traffic distribution with Atlassian Data Center

Monitoring and maintenance

To ensure the ongoing health of your Jira Data Center cluster:

1. Set up regular monitoring of key performance indicators across all nodes.

2. Create alerts for when resource utilization approaches threshold limits on any node.

3. Regularly review and update your node configuration as your Jira usage evolves.

4. Conduct periodic load tests to verify your node configuration can handle peak loads.

Additional resources

• Jira Data Center cluster monitoring

• Data Center infrastructure recommendationsData Center infrastructure recommendations

By carefully managing your node count and configuration, you can ensure that your Jira Data Center instance maintains optimal performance, scalability, and reliability as your organization's needs evolve.