New Relic Insights lets you create clear and easy to use dashboards by writing NRQL query which is similar to SQL. The dashboards could include various type of widgets from new mobile users to website performance. These stats are extremely beneficial as it helps you to provide a better product.
In this tutorial, I’m going to go through some of the NRQL query examples.
New Relic Insights query examples
Finding out the average page duration over a period of a month for mobile users in selected countries
SELECT average(duration) from PageView SINCE last month facet countryCode, pageUrl where countryCode IN ('GB', 'US') and deviceType = 'Mobile'
With mobile users on the rise, this query can help you identify the slow pages of mobile devices on selected countries.
Finding out the average connection setup duration
select average(connectionSetupDuration) from PageView FACET countryCode
This query helps you to identify the average connection setup grouped by country. Website speed is critical these days. You could modify this query to show the speed for only selected countries by using the “WHERE” condition.
Total number of sessions grouped by country
SELECT count(session) FROM PageView FACET countryCode SINCE 4 weeks ago
If you are receiving traffic from multiple countries, it would be good to have an overview of the sessions.
Finding out the most visited pages grouped by the country code.
SELECT count(pageUrl) FROM PageView FACET pageUrl, countryCode SINCE last week
To exclude certain pages you could use the “not like ‘%%’ where the condition
Example: we want to exclude the pages that have “login” in the URL
SELECT count(pageUrl) FROM PageView FACET pageUrl, countryCode SINCE last week Where pageUrl not like ‘%login%’
Finding out the number of sessions by Platform
SELECT count(session) from PageView FACET userAgentOS, userAgentName since 4 weeks ago
You could use stats to have Selenium testing performed on the most used platform.
Finding out how frequent it takes to load pages within X timeframe?
select histogram(duration, 10, 20) from PageView since 3 month ago
The histogram method takes three arguments.
- Attribute name
- The maximum value of the sample range (in our example it is 10 seconds)
- Total number of buckets
We could come to know that 8,161 users have their page loaded between 0.0 to 0.5 seconds. This is great! but we would need to look into bringing more users within this range. The next step would be finding out the which countries our website speed is the slowest and look into ways to improve it.
Finding out new unique users compared with last week
SELECT uniqueCount(session) FROM PageView SINCE 1 week ago COMPARE WITH 1 week AGO
Finding out the number of sessions by device type
SELECT count(session) from PageView facet deviceType
With the queries above you could create your own customized dashboard. Having a dashboard gives you the flexibility to filter out results as well.
My dashboard looks like the one below. The most important widget to me is the number of new unique users.