After deploying a new version of our application server we noted a strong increase in response times of the service.
Preliminary analyses in NewRelic showed that most of the time is spend in the 'mongodb' section of the route. Please have a look at the attached screenshots.
Checking in MongoDB Cloud Manager, we couldn't see any increase in queries or query performance to the cluster.
The only MongoDB related change was a minor bump of the MongoDB Node.js native driver from 2.2.6 to 2.2.8. After pinning the driver version to 2.2.6 the situation normalized again and we are back to the usual performance, as you also can see in the attached screenshots.
Please let me know if I can help with any further detail.