MongoDB Monitoring
and Performance Management

• Monitor whether users are able to connect to the MongoDB instance and execute queries.
• Detect any increase in connection time or query execution time.
• Track connection traffic to each MongoDB instance. Get proactive alerts to any unusual traffic activity via auto-baselining.
• Detect potential connection leaks to the MongoDB instance and identify the affected applications.
• Track connection usage vs. capacity and determine times when the database instance is overloaded.

• Evaluate the memory allocated to an instance and the memory currently in use to identify whether there are times when the memory allocated to a MongoDB instance is insufficient.
• MongoDB uses both the storage engine's cache and the filesystem cache. Monitor the cache usage, dirty cache used, and the number of pages evicted from the cache to identify whether there are any bottlenecks.
• Get alerts when dirty data cache size grows, which could highlight that data is not being written to the disk fast enough. This, in turn, could lead to newer pages being read directly from the disk, which could be detrimental to performance.
• Cursors are used to access data sets. Detect situations when the number of open cursors is increasing. This could indicate that some queries are running for a long time, possibly because they are not using the right indexes.
• Track timed-out cursors, which are indicative of applications that have failed, to close cursors after use.

• Monitor the data size of all collections, index size, and storage size to identify abnormal data growth and plan disk space requirements in advance.
• Monitor the collections in a database that are large. Get detailed information on large collections and optimize collection sizes or consider sharding the collections.
• Track database throughput over time, including read and write rates.
• Monitor queued read and write requests and compare with throughput to determine whether the database server can cope up with the incoming workload.

• Determine whether database replication is working properly on each node and whether each member node of a replica set is running.
• Get alerts on data sync issues between primary and secondary nodes during replication.
• Assess the replication lag - i.e., how far a secondary node is behind the primary node and get alerts if the lag is too high.
• Monitor the oplog window, which is the interval of time between the oldest and the latest entries in the oplog (MongoDB's operations log).

eG Enterprise application performance monitoring allows MongoDB performance to be monitored in the context of the applications using the database server.

• Without requiring any agents for the MongoDB instances, eG Enterprise APM traces all application accesses and reports on slow queries and exceptions during database processing.
• When a specific web page or URL is slow, you don't have to wonder any more as to what is causing the slowdown. eG Enterprise APM gives you the answers at your fingertips.

This eliminates finger-pointing between application development, application operations, and database admin teams.