Please use the form below to provide your feedback. Because your feedback is valuable to us, the information you submit in this form is recorded in our issue tracking system (JIRA), which is publicly available. You can track the status of your feedback using the ticket number displayed in the dialog once you submit the form.

View Metrics for a Capella Columnar Cluster

  • Capella Columnar
  • concept
    +
    Capella Columnar provides metrics that allow you to monitor your clusters.

    You can use the Capella Columnar Monitoring dashboard to access a customizable collection of metrics and gain insight into the performance of your cluster.

    With the Monitoring dashboard, you can:

    • Investigate issues and identify runtime outliers

    • Customize the dashboard to only track specific metrics

    • View all of those metrics in one place

    The Monitoring dashboard includes the following elements:

    Element Description

    Metrics

    The metrics list includes every metric available for Columnar clusters. You can select metrics to add or remove them from your dashboard.

    Timeframe

    The timeframe filter lets you switch between preselected timeframes or set your own custom timeframe.

    By default, the Auto refresh option is turned off. When turned on, it refreshes the dashboard with the most recent data based on the the 1-minute or 2-minute refresh rate you choose. The auto-refresh feature is only available with the 30-minute, 1-hour, or 2-hour timeframe options.

    Metric tile

    The tiles that populate when you choose metrics from the metrics list.

    Monitor Metrics in the Monitoring Dashboard

    To monitor your metrics in the Monitoring dashboard:

    1. In your Columnar cluster, go to Monitoring.

    2. Select the metric you want to monitor from the Metrics drop-down. For a list of all metrics available, see available metrics.

    3. Select a timeframe to apply to the metric. You can choose one of the predetermined timeframes, or click to set a custom timeframe. The dashboard then populates a metric tile with information about the metric and timeframe you selected.

    After you have populated the dashboard with a metric tile, you can view details about that metric:

    • Hold the mouse pointer over a specific part of the metric chart to display information about its date, time, and resources.

    • Hold and drag the mouse pointer over a timeframe on the chart to zoom in on details about that specific timeframe. Doing this updates the timeframe for all metric tiles on your dashboard.

    Capella Columnar draws metrics charts down to 60-second (1-minute) resolution. This resolution is available for charts using the 30m, 1h, and 2h timeframes.

    When you select a longer timeframe, Columnar automatically updates the chart data resolution to the following:

    • 1d down to 360 seconds (6 minutes) resolution

    • 2d down to 720 seconds (12 minutes) resolution

    • 7d down to 2520 seconds (42 minutes) resolution

    • 30d down to 10800 seconds (3 hours) resolution

    Available Metrics

    The following metrics are available in the Capella Columnar Monitoring dashboard:

    Metric Description

    Connections

    The number of connections to the cluster.

    CPU Utilization by Node

    The CPU utilization % by node.

    Ingested Bytes Rate by Link/Bucket

    The per second rate of incoming bytes ingested by Analytics, averaged over the last five minutes by link and bucket.

    Total Ingested Bytes Rate by Link/Bucket

    The total number of incoming bytes ingested by Analytics by link and bucket.

    Linked Ops Rate by Link/Bucket

    The per second rate of linked record operations processed by Analytics, averaged over the last five minutes by link and bucket.

    Total Linked Ops Rate by Link/Bucket

    The total number of linked record operations processed by Analytics by link and bucket.

    Memory Available by Node

    The amount of memory available by node.

    Parse Failure Rate by Link/Bucket

    The per second rate of record parsing failures from linked items, averaged over the last five minutes by link and bucket.

    Total Parse Failure Rate by Link/Bucket

    The total number of record parsing failures from linked items.

    Read Rate by Node

    The per second rate at which disk bytes are read for the Analytics Service, averaged over the last five minutes by node.

    Write Rate by Node

    The per second rate at which disk bytes are written for the Analytics Service, averaged over the last five minutes by node.

    System Load per Node

    The system workload by node.

    Total Requests per Second

    The total number of received Analytics requests for the entire cluster over the last five minutes.

    Total Requests per Second by Node

    The total number of received Analytics requests by node over the last five minutes.

    See Also