If you are the MongoDB administrator, then you need to be able to identify the potential impact of client application changes.
- Keeping historical monitoring data.
- Watching out for spikes in connections.
- Watching out for spikes in operations/second.
- Spotting long lasting, non-indexed queries.
Use MongoDB Atlas to view historical monitoring data diagnose all spikes.
Connection spikes may caused by:
- Exhausting connection pools.
- Incorrect connections to production environment.
- Analytical workload and reporting tools.