Achieving optimal performance from your Filr deployment requires that you do the following:
Assess system needs and plan to meet those needs as outlined in Section 3.0, Planning a Large, Expandable (Clustered) Deployment.
Deploy Filr according to your plan by following the basic order outlined in Section 4.0, Setting Up a Large, Expandable (Clustered) Deployment.
Adjust and tune system settings and resources on an ongoing basis as workload details become clearer and Filr usage increases.
This section is a companion to the Filr 1.1 white papers that are published on the online documentation website: Novell Filr 1.1: Test Report—Net Folder Synchronization (OES)
and Novell Filr 1.1: Test Report—Workload Scalability/Capacity (OES)
. The white papers present and explain performance, scale, capacity, and other test results; this section offers suggestions on related topics at a more general level.
Ultimately, each Filr deployment is unique. Therefore, we can only offer the general guidelines in the sections that follow.
Section 5.1, Ensure That Hardware Recommendations Are Met or Exceeded
Section 5.2, Ensure Optimal Network and Disk Bandwidth for the /vashare Mount Point
Section 5.5, Dedicate the Third Filr Appliance to Synchronization
Section 5.7, Understand How Data and Synchronization Affect Performance
Section 5.8, Bring Net Folder and Home Directory Users Online in Phases
Section 5.9, Rely on Ganglia for Help with Optimizing RAM and CPU Allocation