From the beginning, 10gen has focused MongoDB on four key areas: flexibility, power, speed, and ease of use. In terms of ease of use, we are very much interested not only in improving usability for developers, but also for IT operations. Early in MongoDB’s development, we released MMS (MongoDB Monitoring Service) to offer users visibility into the right metrics to manage and optimize applications during development and in production. To continue to improve our users’ ability to manage MongoDB in production, there is a powerful new type of alert type available in MMS - Metric Min/Max Value.
The Metric Min/Max Value alerts provides alerting on a variety of host types and corresponding MongoDB performance metrics. For example, you can say ...Alert me if any of my secondaries experience a repl lag of greater than <x> mins.â€œ
This alert type is now flexible enough to provide alerts for the most important performance boundaries that's specific to your application's performance profile.
What does each new alert mean?
You can alert against a number of different host types listed below.
With a replication-enabled host type selected, you'll also have the option to select a specific replica set for this alert, or to have the alert apply to all replica sets.
There's a wide variety of performance metrics to choose from for your alerts. To enumerate all the metric types would be intense. The options are essentially straight from existing MMS chart types, and hopefully they're pretty self explanatory. Example of available metrics when ...Secondariesâ€œ host type is selected:
Limitations: no hardware stats
For now, even if you have hardware stats configured and enabled, they cannot be used for Metric Min/Max Value alerts.