Docs » View Usage Information (DPM plans only)

View Usage Information (DPM plans only) πŸ”—

Note

The information in this document applies to organizations whose subscription plan is based on the rate at which you are sending datapoints to SignalFx (DPM). If your organization’s usage is based on the number of hosts or metrics that SignalFx is monitoring for you, see View billing and usage information.

Admin users in your SignalFx organization can view the usage information for the organization as a whole.

About usage reports πŸ”—

SignalFx provides summary and detail to help you understand and manage your metrics submission.

  • The first, Datapoints per minute detail report, shows the total number of datapoints sent for each minute of a month (or month to date, for the current month). This information is helpful for understanding whether your rate of data submission has stayed within the limits of your SignalFx subscription.
  • The second, Datapoints per dimension report, shows on a daily basis the number of datapoints and time series for a given dimension value, as well as its average reporting frequency (resolution). It is useful for understanding the nature of the data your organization is sending so you can tune it accordingly. For example, you might notice that there are some metrics which you do not want to send at all, and conversely, you might see that there are some metrics that you want to send at a higher resolution.

Using the billing and usage page πŸ”—

To view available usage reports, open the Settings menu at far right on the navigation bar, hover over Organization Settings, and then select Billing & Usage. You see a usage chart that shows average and maximum DPM for the current billing period. Hovering over different points in the chart shows actual values below the chart. To see average and maximum usage for a previous period, choose the desired period from the dropdown menu.

../_images/dpm-usage-max.png

To view usage reports available for download, click View detailed usage reports. The reports available on the Billed Usage tab will be shown.

../_images/dpm-summary-tab.png

Different reports are available on the Billed Usage and Usage Breakdown tabs. Clicking on the icon for a report will download the report. (In some browsers, you might have to right click on the icon and save the report.) All reports are tab-separated text files. They are designed to be imported into a spreadsheet application, such as Microsoft Excel, Google Sheets or Apple Numbers.

Datapoints per minute detail report πŸ”—

This report, available on the Billed Usage tab, shows the total number of datapoints sent for each minute of a month (or month to date, for the current month). It follows your billing period, and uses the month in which a billing period starts as the label in the report link. For example, if your billing period starts on the 10th of the month, then a link for β€˜March 2018’ will cover the period of March 10 through April 9, 2018.

Report format πŸ”—

The report is a tab-separated text file. It is designed to be imported into a spreadsheet application, such as Microsoft Excel, Google Sheets or Apple Numbers.

Report contents πŸ”—

The report has five columns:

  • Date

    Follows the mm/dd/yy format

  • Time

    24 hour hh:mm UTC

  • Streaming Datapoints

    The number of datapoints received by SignalFx through our ingest APIs for the specified date and time. This includes datapoints sent in using the API directly; via a supported agent like collectd using our write_http plugin configuration; through a client library; or using the SignalFx Gateway (formerly called the metric proxy).

  • Backfill Datapoints

    The number of datapoints received by SignalFx through the backfill API for the specified date and time.

  • Throttled Datapoints

    The number of datapoints actively dropped by SignalFx for the specified date and time.

For past months, the report includes streaming, backfill and throttled datapoints for all of the minutes in that month. For the current month, the report includes the datapoints for the month-to-date, and is updated once per day.

Using the report πŸ”—

The report shows you the detail of how many datapoints SignalFx is receiving from your organization on a per-minute basis. In general, this is useful if you want to determine whether your usage is in line with your subscription plan. Using the data, it is easy for you to calculate your average usage, how many minutes in the month you have been over or under your plan, and by how much.

Aggregate Datapoints per minute detail report πŸ”—

If you are working in a SignalFx organization that has a parent-child relationship with other organizations, you will see an option for the Datapoints per minute report that includes information on child organizations. Like the Datapoints per minute detail report, this report shows the number of datapoints sent to SignalFx for each minute within the month; however, this report also includes this data for all child organizations.

Datapoints per dimension report πŸ”—

This report, available on the Usage Breakdown tab, shows on a daily basis the number of datapoints and time series for a given dimension value, as well as its average reporting frequency. Reports are for the 24-hour period (in UTC) represented by that date.

You can select or type in a date for this report. Data is retained for a limited amount of time.

Report format πŸ”—

The report is a tab-separated text file. It is designed to be imported into a spreadsheet application, such as Microsoft Excel, Google Sheets or Apple Numbers.

Report contents πŸ”—

The report has seven columns:

  • Dimension Name

  • Dimension Value

    Dimension Name and Dimension Value are the key/value pairs of the dimensions that are sent in with your metrics. Unique combinations of dimensions and metrics are represented as time series in SignalFx.

  • Number of Contributing Metric Time Series

    The number of metric time series associated with the dimension name and value.

  • Total Datapoints Received per Day

    The total number of datapoints received by SignalFx during the 24 hour period represented by the date of the report, again as associated with the dimension name and value.

    Note that if you sum up the Total Datapoints Received per Day column, it is possible that your total will be larger than what is reported in the Datapoints per Minute detailed billing report. This scenario would occur if any of your metrics report more than one dimension, in which case that datapoint would be counted multiple times in this report (once for each dimension).

  • Average DPM

    The datapoints per minute, averaged across the number of time series associated with the dimension name and value, and over the course of the 24 hour period represented by the date of the report.

  • Average Reporting Frequency for the Day

    The time (in seconds) between successive datapoints, averaged across the number of time series associated with the dimension name and value, and over the course of the 24 hour period represented by the date of the report. For example, a value of 10 means the data is being sent every 10 seconds, i.e. has a 10s frequency; a value of 60 means that the data is being sent every minute, i.e. has a 1m frequency; and a value of 300 means that the data is being sent every 5 minutes, i.e. has a 5m frequency, as is the case with standard AWS CloudWatch metrics.

    Note that frequency is calculated as an average across all of the time series associated with the relevant dimension value, rather than measured for each individual time series. As a result, it may contain outliers (e.g. a time series that is reporting more slowly or with greater jitter or lag) that skew the average. For example, for data being sent every 5 minutes (300 seconds), you might see a value of 280, or a value of 315. Frequency should be treated as an approximate value that guides what you do with your metrics, rather than a way of auditing the precise timing of them.

  • Number of Contributing Metric Time Series Which Were Created Today

    The number of metric time series associated with the dimension name and value that were created over the course of the 24 hour period represented by the date of the report.

Using the report πŸ”—

  • Metric names

    A good starting point for using the report is the sf_metric dimension. The dimension values are the names of the metrics being sent in, and you can see how much data is being sent in for each metric and at what resolution. Armed with this information, you can decide whether you want to send in that metric to SignalFx, and if so, how often.

  • collectd

    If you are using the SignalFx collectd agent, then another place to gain insight is to look at the plugin dimension. collectd collects metrics via plugins, and plugins are easy to install / uninstall or configure (for their reporting interval).

  • Amazon Web Services

    If you are using the Amazon Web Services integration, you should also look at the namespace dimension. The AWS integration lets you select the services for which you want data, such as AWS Cloudwatch, and the namespace values tell you how much data is being sent in for each AWS service that is turned on.

  • Hosts and instances

    Finally, many people find the host, instanceId or similar dimensions worthwhile, as hosts or instances are often associated with different stages or environments. For example, you might decide that you want to have fewer metrics coming in from your test environment, and more from production. Knowing which hosts map to which stage or environment gives you insight into this question, and lets you manage your data submission accordingly.