New dimensions and more comprehensive insights with "empty" values
We're excited to introduce two new improvements to our Advanced Analytics experience:
More comprehensive insights with "empty"
We now display counts for "empty" values in our analytics experience. "Empty" indicates a request exists, but there is no specific value associated with an attribute of that request (e.g. route). This enhancement allows users to quantify API calls that don't match specific attributes, providing more comprehensive insights.
Key benefits:
Identify the number of API calls that don't match a route
Count requests without an associated consumer
Gain a complete picture of your API traffic, including undefined attributes
This feature empowers users to better understand issues early and fix them if necessary.
Additionally, you are now able to use `Is empty` and `Is not empty` operation inside the Explorer to filter for any requests that match that criteria.
New dimensions
Advanced Analytics users now have access to three new dimensions to Explorer and Reports: "Response Source", "Upstream Status Code", "Upstream Status Code Group", and "Data Plane Version". These new dimensions enhance our users' ability to filter and group data, providing deeper insights into API performance and error sources.
With these additions, users can now easily distinguish between proxy errors in Kong and upstream errors, allowing for more precise troubleshooting and performance analysis. The new dimensions enable users to answer critical questions such as:
- How many proxy errors occurred in Kong during a specific time period?
- How many upstream errors were observed in Kong during a given timeframe?
- What is the distribution of proxy errors versus upstream errors over time?
- Is there any traffic still handled by older versions of Kong Gateway?
Last but not least, we also added the ability for users to visualize the total sum for response and request size. This is now available in the metrics dropdown inside our Explorer interface. This allows users who have metering use cases to track usage based on size.
