Check your query and index performance with the Query Targeting Chart

September 19, 2016

We’re excited to announce a new feature for Monitoring in both Cloud Manager and Atlas: The Query Targeting Chart. This chart tracks two variables, the first is “scanned/returned” and the second is “scanned objects/returned”.

https://webassets.mongodb.com/_com_assets/blog/tblr/66.media.tumblr.com--d1cca82c135ff5d50e4ff6af52a9d15d--tumblr_odt0epEg5y1sdaytmo1_400.png

“Scanned/returned” refers to the ratio between the number of index items scanned and the number of documents returned by queries. If this value is 1.0, then your query scanned exactly as many index items as documents it returned – it’s an efficient query. This is available for MongoDB 2.4 and newer.

“Scanned objects/returned” is similar, except it’s about the number of documents scanned versus the number returned. A large number is a sign that you may need an index on the fields you are querying on. This metric is available for MongoDB 2.6 and newer.

For a little more understanding of this graph, let’s talk about a collection with 1000 documents in it. We then issue a query without an index (so it is a collection scan). Scanned objects/returned for this query could be as bad as 1000, but the average value would be 500. Now, let’s put an index on that same query, return one document and we only have scanned one document. This means that scanned/returned is 1, and scanned objects/returned is also 1. Finally, let’s say you do a covered query, in this case the scanned/returned is 1, but the scanned objects is 0, because the index has all the data you requested, so you didn’t need to query any objects!

This feature is available for all Cloud Manager and Atlas deployments. We believe this new chart will help you refine your queries and indexes to get the best performance out of your MongoDB deployment. However, if you need more help, the Visual Profiler as part of Cloud Manager Premium can help you identify slow queries and suggest indexes as well. Contact your Account Executive for more information about MongoDB subscriptions with access to Cloud Manager Premium.

Peter C. Gravelle is a Technical Account Manager at MongoDB, Inc. He can be found via Atlas’ chat option as well as in tickets. He can also be found in New York City.

Previous Article
Building Applications with MongoDB's Pluggable Storage Engines: Part 2
Building Applications with MongoDB's Pluggable Storage Engines: Part 2

In the previous post, I discussed MongoDB’s pluggable storage engine architecture and characteristics of ea...

Next Article
Building Applications with MongoDB's Pluggable Storage Engines: Part 1
Building Applications with MongoDB's Pluggable Storage Engines: Part 1

This is the first in a two post series about MongoDB’s pluggable storage engines. This post discusses chara...