Fabric Health Analytics BETA

Fabric Health Analytics is delivered as BETA and is subject to change in the upcoming release(s).

Fabric Health Analytics (FHA) in GigaVUE-FM is a standalone service that provides data visualization capabilities for the following entities in GigaVUE-FM:

  • Physical resources, specifically the nodes and the ports
  • Virtual resources
  • Alarm Management services
  • GigaVUE-FM CPU, Memory and Disk Storage services

Using FHA1 you can create visual elements such as charts that are embedded as visualizations. The visualizations are grouped together in dashboards. You can also create search objects using FHA. Dashboards, Visualizations and Search Objects are called FHA objects.

Fabric Health Analytics provides the following advantages:

  • Real-time data for visualization as the required data is taken from GigaVUE-FM.
  • Data to be analyzed and visualized is fetched based on the access control rights of the user.
  • Trending Analysis: Visualize the trends in the traffic using pre-defined widgets.
  • Capacity Planning: Utilization of resources based on health and inventory summary data.
  • Generation of reports based on the available data.

Rules, Notes, and Limitations for Fabric Health Analytics

  • All GigaVUE-FM users can create, edit and delete FHA objects2. However, you can perform these operations only on the FHA objects created by you. You cannot delete or edit system-defined objects such as dashboards and visualizations.
  • There is no limit on the number of dashboards per user.
  • You can only view the FHA objects created by other users, but you cannot edit them.
  • The data available in dashboard is controlled by Role Based Access Control. That is, the data fetched depends on the accessibility rights of the user based on the user role and user-defined tags.
  • The GigaVUE‑FM backup/restore operation preserves both default and custom dashboards and visualization. This allows the dashboards and visualizations created in the earlier software versions to be restored in software version 5.13.00. This enables rapid resumption of GigaVUE‑FM services and also provides the users the ability to operate and monitor the visualizations in GigaVUE‑FM.
  • Statistical dashboards display data in the following two tabs.
    • Metric: Displays maximum and average values of statistical counters for a specific time period, example Rx Ports Aggregated Max Traffic, Rx Ports Aggregated Average Drop Traffic
    • Trends: Displays trend of statistical counters for a specific time period, example Rx Port Statistics (Bits/s), Tx Port Statistics (Bits/s), Tool GigaStream Maximum Rx Rate(pps) by member port Alias
  • The Inbound and Outbound Port Statistics dashboards display metric values for the top 1000 port elements in visualizations (as the total count it set to 1000).
  • By default, all statistical dashboards display data based on a pre-selected cluster ID to avoid performance issues in a scaled environment.
  • For rate-based visualizations in the default system dashboards and cloned system dashboards, the axis-min setting (under panel option) is set to zero (0). An empty graph is therefore displayed when the data point values are zero. However, for visualizations in the dashboards created using the Create Dashboard option, you must manually change the axis-min setting to '0' for an empty graph to be displayed.
  • For gauge-based visualizations "no data to display" message is received in case of the following scenarios:
    • GigaVUE-FM or the device is down during a particular time interval.
    • No traffic in the device and the axis-min panel option is set to value > 0.
  • For rate-based visualizations, "no data to display" message is received in case of the following scenarios:
    • With Drop Last Bucket set to 'Yes' and time interval range < 15 minutes or time range has less than three data points.
    • With Drop Last Bucket set to 'No' and time interval range < 10 minutes or time range has less than two data points.

GigaVUE-FM Statistics and Data Roll-up

Until software version 6.0.00, the statistical dashboards of the physical and logical resources display data collected for a period of 30 days with 5 minute granularity (considered as the base index). Starting from software version 6.1.00, the statistical dashboards display data collected for a period of 35 days.

Resource

Related Statistical Dashboards
Ports
  • Inbound Port Statistics
  • Outbound Port Statistics
Maps Map Statistics
Nodes Inventory Statistics
GigaStreams
  • Circuit and Stack GigaStream Statistics
  • Tool GigaStream Statistics
GigaSMART GS Group Statistics

With Data Rollup, GigaVUE-FM allows you to collect, summarize and display historical data up to 120 days with hourly granularity. Refer to the following notes:

  • Starting from software version 6.1.00, Rollup is enabled by default. To disable Data Rollup, navigate to the Storage Management page. Refer to the GigaVUE Administration Guide for details.
  • For 6.1.00 release, Rollup is applicable only for ports. You can search and analyze port statistics for a duration of up to 120 days. GigaVUE-FM collects statistical data from the day you upgrade to software version 6.1.00. Complete 120 days of data will be visible only from 121st day (and not backwards from the day GigaVUE-FM was upgraded).
  • On upgrading to software version 6.1.00, ensure to have increased disk space for data roll-up. Refer to the Virtual Computing Resource Requirement in Scaled Environments section in the GigaVUE-FM Installation Guide for detailed information.

Control Filters in FHA Dashboards

You can use control filters in your dashboards to filter and display the data you want to explore:

  • To filter data in the visualizations based on the required tag keys and tag values, clone the required statistical dashboard and edit the Sample - Tags option in the Control Filters. Refer to Filter Data Using Tags in Control Filters
  • Starting from software version 6.2.00, Cluster ID is a mandatory control filter field in the statistical dashboards. When cloning the dashboards and visualizations, you must use at least one of the following control filter fields to filter the data:
    • Cluster ID
    • Tag

Reference Topics

Refer to the following sections for details: