You are here:
Limits and Limitations for Trending Data in Analytics
Trended datasets count toward the overall CRM Analytics platform limits, including total number of rows. To prevent the performance of your org from degrading, certain limitations are placed on snapshot size and the number of trended datasets each user can create.
Check out the following limits on trending data.
To control the impact that trended datasets have on general org-wide limits, configure these limits, available on the Analytics Settings page:
- Max number of rows per snapshot is an option to control the maximum number of rows that a trended report can have when a snapshot is taken. If the report exceeds the limit, the snapshot isn’t taken and an error email is sent to the user.
- Max total number of rows in a trended dataset is an option to limit the maximum number of rows that the Analytics dataset containing the snapshots can have. If the dataset exceeds the limit, the oldest snapshot is deleted from the dataset.
- Enable automatic deletion of inactive datasets is an option that is on by default. Trending dashboards that aren't being updated or accessed for the number of days configured in Number of days to keep inactive datasets are deleted. The default number of days a trending dashboard can remain inactive is 100. The dashboard owner receives an email warning a week before deletion. When the dashboard and dataset are deleted, the owner receives an email confirmation.
Roll out the capability gradually to your user base and observe data usage and performance. Stop trending datasets that are no longer needed for active use and delete the underlying trended datasets to recover storage limits.
The best candidates for trending are reports that can be visualized in time and have some frequently changing values. Examples: open opportunities pipeline or escalated cases.
You can’t use trended datasets as sources in recipes. To prepare data from trended datasets, use a dataflow.
Some situations can prevent a report that is trended in Analytics from successfully capturing the data for a particular run. These situations include, but are not limited to: Jobs involving the User object, site switching, and instance maintenance. If a trended report attempts to run under these conditions, it fails and the dataset does not have records added for that run. You can’t reschedule or run a trending job again; however, future runs that are not subject to these conditions process as normal.
Users whose profiles have View Encrypted Data enabled can't create trended datasets because Analytics does not support encrypted data.
Only the user who starts trending the report can stop trending it. The user can share the trending dashboard with other users. If another user wants to trend the same report, the user must create a trend, which does not affect other trends for the report.
After trending begins, do not change access rights to the underlying report or add or delete dimensions and measures. Changing the report or the access rights can cause an error or inaccuracies at the time a snapshot is taken.
If you modify a field in a report that is trended, the change is reflected in the associated dashboard from the time of the change onward. If you change other aspects of the report, such as filters, those changes aren’t reflected in the associated dashboards.
Snapshots are taken every Sunday at midnight (local time), unless you set a different schedule when trending a report. Processing and surfacing the data requires more time, which can delay snapshots.
If trending is stopped and then restarted, a new snapshot dataset is created, with a new trending dashboard. The previously trended dataset isn’t included in the new dashboard. Data can be moved into the new dataset using the Salesforce data APIs.
Trended datasets are created in their own private folders. You can use the Update Dataset button to move the dataset from a private to a public folder.
The dataflow that creates a trended dataset is auto generated and can’t be viewed or modified. This restriction is required because the underlying trending logic can be sensitive to changes in the dataflow.
You can’t customize the XMD of trended datasets. Customize Actions functionality aren't available for trended datasets.
The dashboard designer is enabled by default for most orgs. Make sure that it remains enabled for your org. To render, trending dashboards rely on the dashboard designer being enabled.
Boolean fields in a trended report are treated both as a measure (for calculations) and a dimension (for sorting).
If you modify the dashboard for a trended report, the changes are reflected in the next trend update.
