API audit

Important:

This feature is currently available as a preview.

This displays the number of calls made to GetAll - Jobs and QueueItems endpoints over the last 30 days, with the option to drill down into a particular day.

The graphs display a green bar if our 100 requests/minute/tenant rate limit is observed, and an orange bar if it is exceeded.

Each graph includes a View API button which directs you to the corresponding Swagger endpoint.

This requires the Audit - View permission.

Important notes:

• The count starts after the first 10 API calls.
• The data is displayed with a delay of maximum 20 minutes.
• Some internal calls may be included in the graphs, but they are not counted towards the rate limit. You may, however, see internal calls exceeding the limit, but they will have a success status code.
• Graph times are expressed in UTC.

Spike details

You can view spike details for each minute logged in the API audit panel.

To do that, select the spike corresponding to the desired minute. This opens a side panel which lists a maximum of 100 results with the most requests, and which includes information such as the process, the folder where the process resides, the host name, the number of calls performed to that endpoint during the selected minute, the status code, and the user agent.

If the API call originates from an external source or if we do not have enough details about the process, the corresponding columns display N/A.

These are the actions that you can perform within the selected spike details side panel:

• Click the expand button in the top right corner to open the panel in a full-sized window.
• Click the eye icon corresponding to a process to see the details of the associated job.
  Note: This option is not available if you do not have view permissions for the folder where the process is configured.

Important:

The number of requests is counted starting with the very moment that a request is made, down to the exact second. However, for simplicity, our charts display one-minute increments instead of second increments.

This means that you can have the following scenario:

• You make 50 requests between 1:10:30 PM and 1:10:59:99 PM.

• You then make another 60 requests between 1:11:00 PM and 1:11:30 PM.

The chart bars corresponding to each minute will be green, but, in reality, the extra 10 requests will be rate limited.