Versions Compared


  • This line was added.
  • This line was removed.
  • Formatting was changed.
Comment: Published by Scroll Versions from space ML1 and version 9.0

This section describes the potential limiting platform factors on cluster performance, how to measure whether a cluster is approaching or exceeding those limits, and options available to remedy such conditions. Note that by "platform Platform factors," we are referring refer to hardware resources such as CPU, memory, and disk, and network I/O subsystems. This section does not attempt to address the matter of Optimizing Performance Based on Query Execution PlansFor potential software-related factors, please see Managing Data Distribution, or  Load Balancing ClustrixDB with HAProxy, each covered elsewhere. or Understanding the ClustrixDB Explain Output.

Table of Contents

CPU Load


There are other cases where CPU contention becomes a bottleneck even though the cluster is not being fully utilized; that is, load is not optimally balanced across the cluster. This can be due to external factors such as an inefficient query, or client connections being poorly distributed across nodes (if not connecting through the VIP). A suboptimal configuration could also be a culprit, such as having a table that is not distributed evenly across the cluster, although the system goes to great lengths to automatically manage this.  .

Monitoring CPU Load

CPU load reflects the busyness of each node’s CPU cores in the ClustrixDB cluster.