Performance monitoring using IBM Spectrum Scale GUI

The IBM Spectrum Scale™ GUI provides a graphical representation of the status and historical trends of the key performance indicators. This helps the users to make decisions easily without wasting time.

The following table lists the performance monitoring options that are available in the IBM Spectrum Scale GUI.
Table 1. Performance monitoring options available in IBM Spectrum Scale GUI
Option Function
Monitoring > Statistics

Displays performance of system resources and file and object storage in various performance charts. You can select the required charts and monitor the performance based on the filter criteria.

The pre-defined performance widgets and metrics help in investigating every node or any particular node that is collecting the metrics.

Monitoring > Dashboards Provides an easy to read and real-time user interface that shows a graphical representation of the status and historical trends of key performance indicators. This helps the users to make decisions easily without wasting time.
Monitoring > Nodes Provides an easy way to monitor the performance, health status, and configuration aspects of all available nodes in the IBM Spectrum Scale cluster.
Files > File Systems Provides a detailed view of the performance and health aspects of file systems.
Files > Filesets Provides a detailed view of the fileset performance.
Storage > Pools Provides a detailed view of the performance and health aspects of storage pools.
Storage > NSDs Provides a detailed view of the performance and health aspects of individual NSDs.
Files > Transparent Cloud Tiering Provides insight into health, performance and configuration of the transparent cloud tiering service.
Files > Active File Management Provides a detailed view of the configuration, performance, and health status of AFM cache relationship, AFM disaster recovery (AFMDR) relationship, and gateway nodes.

The Statistics page is used for selecting the attributes based on which the performance of the system needs to be monitored and comparing the performance based on the selected metrics. You can also mark charts as favorite charts and these charts become available for selection when you add widgets in the dashboard. You can display only two charts at a time in the Statistics page.

Favorite charts that are defined in the Statistics page and the predefined charts are available for selection in the Dashboard.

You can configure the system to monitor the performance of the following functional areas in the system:
  • Network
  • System resources
  • NSD server
  • IBM Spectrum Scale client
  • NFS
  • SMB
  • Object
  • CTDB
  • Transparent cloud tiering. This option is available only when the cluster is configured to work with the transparent cloud tiering service.
  • Waiters
  • AFM
Note: The functional areas such as NFS, SMB, Object, CTDB, and Transparent cloud tiering are available only if the feature is enabled in the system.
The performance and capacity data are collected with the help of the following two components:
  • Sensor: The sensors are placed on all the nodes and they share the data with the collector. The sensors run on any node that is required to collect metrics. Sensors are started by default only on the protocol nodes.
  • Collector: Collects data from the sensors. The metric collector runs on a single node and gathers metrics from all the nodes that are running the associated sensors. The metrics are stored in a database on the collector node. The collector ensures aggregation of data once data gets older. The collector can run on any node in the system. By default, the collector runs on the management node. You can configure multiple collectors in the system. To configure performance monitoring through GUI, it is mandatory to configure a collector on each GUI node.
The following picture provides a graphical representation of the performance monitoring configuration for GUI.
Figure 1. Performance monitoring configuration for GUI
Performance monitoring configuration for GUI

The mmperfmon command can be used to query performance data through CLI, and configure the performance data collection. The GUI displays a subset of the available metrics.