Get comprehensive insights into the internet’s middle mile.
Akamai DataStream (now in beta) captures interactions between end users, Akamai edge servers, and origin servers (collectively known as the “middle mile”) as logs; these logs are then aggregated and made available to developers through push and pull APIs.
These logs include data and insights on content delivery network (CDN) health, latency, offload, events, and errors, all of which gives you near real-time, detailed visibility into how your CDN is performing. This, in turn, helps you accelerate the speed of development and reduce the time required for troubleshooting and maintenance.
Watch this quick 90-second video that provides an overview of what DataStream can do for you:
- Near real-time visibility: Access a wide range of metrics to help you understand your CDN’s health and also to help you understand exactly what’s impacting your middle-mile performance.
- Data on demand: Data is available to you through push and pull APIs and retained for up to 24 hours.
- Granular control and selectability: You have the ability to select specific log fields in order to focus on the metrics that matter most to you.
- Pre-aggregated metrics: You can choose to receive pre-aggregated metrics over a chosen window of time. Unlike raw logs, this is meaningful, actionable data without requiring significant downstream processing at your end.
- Seamless integrations: DataStream supports out-of-the-box integrations with Amazon S3, Splunk, and Sumo Logic.
How DataStream works
For every transaction (i.e., request-response cycle) on the Akamai platform, DataStream tells Akamai edge servers to send the logs for those transactions to a backend aggregator; after that, the logs are available to your operations team via REST APIs for up to 24 hours. You can use push APIs to pipe the logs to your logging environments or pull APIs to get on-demand data in near real time.
DataStream aggregation is based on a tumbling window over the selected aggregation time frame. Tumbling windows are a series of fixed-sized, non-overlapping, and contiguous time intervals.
Here’s an example of a 10-second tumbling window, in three intervals:
DataStream aggregation operators start collecting live data based on your chosen aggregation time frame at the time of stream definition. For example, if a stream was created at 10:00 with a selected time frame of 5 minutes, DataStream will collect the data from 10:00 to 10:05 and then run an aggregation on this 5 minutes of data.
When setting up your aggregation, you can select from a range of different time frames, as seen here:
DataStream supports all Akamai content delivery products and integrates with top third-party data-analytics tools like Sumo Logic, Splunk, and Amazon S3 buckets so you can visualize and retain your DataStream logs for longer-term storage and analysis.