Observe and Optimize a CDN Distribution

Figure: Distribution Monitoring View

The Dashboard view offers the following:

• Time series view for statistics such as Requests, Data Transfer, and Bandwidth. This is shown in a graph where hits and misses are displayed. Hover over the graph to see specific quantity information for that time point.

• Donut chart for cache hit versus miss ratio.

• HTTP Status codes show the number of response codes in each of the categories. Hover over the horizontal bar to see the values expressed as percentages.

• Latency shows the time taken between requests and responses for both hits and misses.

• HTTP version metrics.

• TLS version metrics.

• Client requests by country. Hover over a country to see the number of requests originating from that country.

• Top 5 countries providing requests, shown as a bar chart. Hover over a bar to get the specific value for that country.

Figure: Distribution Requests Monitoring View

The Requests tab provides greater insight into requests. There is a bar chart as well as a list of the incoming requests. Use the time period drop-down menu and refresh button to set the time constraint and update the data shown.

Interact with the bar chart as follows:

• The bar chart at the top shows the number of requests and response codes in each time period.

• Use the Add Filter to see only certain request based on your filter selection.

• Use the colored check-boxes below Add Filter to quickly filter out specific response types.

• Hover over a bar to get specific bar values (request counts) for that time period.

• Click and drag within the bar chart to zoom into a range of time. Use the time period drop-down menu to zoom back out.

The list view below the chart shows specific information for individual requests. The time period and filters also affect the requests shown in the list view.

Interact with the list view as follows:

• Use the Search field to only show list entries that match your search criteria. The search will work for all items in the list, not just the items on the current page; however, the search will not find information in columns that are not displayed (see the gear icon description below).

• Click the gear icon ( ⚙ ) to select which columns are shown in the list view. Time and Client IP will always be shown.

• Click the arrow ( > ) at the beginning of the row to see the specific request in JSON or YAML format.

Figure: Distribution Monitoring Requests

You can also click the blue Forensics side tab to view the Forensics side panel. This is an easier way to filter your request to zero in on specific issues. For instance, you can filter by site and by top request page simply by selecting check boxes in the Forensics panel. This will create a search query for the filter at the top of the window.

Figure: Distribution Monitoring Requests Using Forensics

Click the Alerts tab to load the alerts view. The active alerts are displayed by default.

You can filter the display for alerts of a specific severity using the severity selection options. All severity types are selected by default. Select a severity selection option to hide the alerts for that severity. You can again select it to display alerts for that severity.

Note: Severity selection options are color-coded and located above the Add filter option.

Use the toggle selector and click All Alerts. The All Alerts view shows graph for alerts over a specific period. The list of alerts is displayed beneath the graph.

Hover mouse pointer over a graph bar to view the alerts information specific to the time interval in which the bar is generated. Selecting the bar updates the graph and the list beneath the graph for the interval in which the bar is generated.

Figure: CDN Alerts

Click > for any alert entry to display its details in JSON format.

Figure: Distribution Cache Purges

The Cache Purges tab show past purges for your distribution in a table. Each row in the table provides a unique ID for the purge, the version and information on the date/time of the purge. To see details of the purge, click ... > Show Status in the Actions column for the purge you want to view.

Figure: Distribution Cache Purge Status

This displays a table showing the sites that were purged and some overview information about each site's purge. For details on the site purge, click the ID in the Service OP ID column for the site.

Figure: Distribution Cache Purge Site Details

Click Back to view more site purges or click Cancel and Exit to return to the Cache Purges page.

You can also deploy a cache purge from here by following these steps:

• Click Add Cache Purge.

  

  Figure: Distribution Cache Purge Request

• Select what to purge from the Purge Options drop-down list.

  • All: Your entire cache will be purged.
  • URL: Purge that portion of your cache identified by the entered URL.
  • Hostname: Purge all content for the entered hostname.
  • Pattern: Purge all content that matches the entered pattern. The string should be entered in Regex form. For example, "images/.*.(png|jpg)" will match PNG and JPEG files in the images directory.

• Select an option from the Purge Type drop-down menu to specify how the purge is to operate.

  • Soft Purge invalidates the cache entries, which means the content will be replaced on the next request if the content is stale.
  • Hard Purge removes the content from the distribution and forcing the next request to retrieve the content from the origin server.

• Click Save and Exit to deploy the purge.

This will create a new entry in the Cache Purges table providing information on your purge.

Note: You can also deploy a purge from the Distributions page. See Purge Content in Cache above.