Shared disk write latency measures the time taken to write data to your shared home directory, which serves as common storage across all nodes in your Data Center cluster. This metric is crucial for maintaining optimal system performance and user experience.

How does high write latency affect performance?

High shared disk write latency significantly impacts two critical areas of your instance:

• Content Operations
  Users experience slower attachment handling and delayed space operations, including sluggish uploads, downloads, and exports. System tasks like profile picture updates and index rebuilds take longer to complete, directly affecting user productivity.

• System Stability
  Extended write times increase load on application servers and risk incomplete operations. Your cluster may experience delayed synchronization between nodes, potentially leading to timeout issues during file operations.

What's the recommendation?

Verify network connectivity

Network performance often represents the primary bottleneck for shared storage operations:

Basic connectivity tests

1. Run these commands from each node to check connectivity:

   
   

   ping -c 10 nfs-server-address netstat -an | grep nfs-server-address

   
   

   
   

2. Check network stability:

   • Go to Administration > System Information

   • Review network statistics

   • Monitor connection status

3. Common network issues:

   • High round-trip times (>1ms)

   • Packet loss or retransmissions

   • Unstable connections

   • Network interface errors

Review your network configuration to ensure:
• Optimal network topology for your deployment
• Sufficient bandwidth for your usage patterns
• Properly configured network cards and switches
• Efficient routing between nodes and storage

Learn about network diagnostics

Review storage configuration

Examine your storage setup to identify and resolve bottlenecks:

Storage setup

1. Check configuration:

   • Review mount options

   • Verify storage tier assignments

   • Examine filesystem settings

2. Monitor competing processes:

   • Backup operations

   • Antivirus scanning

   • System maintenance

   • Security operations

Implement storage best practices:

• Use recommended filesystem types for your deployment

• Configure RAID settings appropriate for your usage

• Apply optimal mount options for your environment

• Coordinate windows backup and optimize antivirus exclusions

More on managing Confluence data

Monitor system resources

Implement monitoring, focusing on key metrics that affect shared storage performance:

• Track system resources including CPU utilization, memory usage, and file handle consumption
• Monitor storage activity patterns, particularly during peak usage and backup windows
• Document baseline performance and analyze trends to identify potential issues early

Set up alerts for:

• Write latency exceeding 20ms

• Unusual spikes in concurrent operations

• File handle usage approaching limits

• Backup operation delays

Learn more about live monitoring using JMX interface