Unexpected Surge in MongoDB Shared Cluster Connections; Concerns About Debugging and Reliability

During Friday 5th of May around 4pm UTC, my MongoDB M0 shared cluster’s connections randomly went from ~6 to 500 connections, despite there being no activity on the app. The database completely stopped working during that time, I couldn’t connect to it or even view the collections in Atlas UI. I turned off the app server and DataGrip on my machine, waited one hour and by then the connections had went to 0 again. This app has been running for months without this ever happening before so this was definitely out of the ordinary.

I am concerned because I was not able to debug from where those connections were coming from, as logs do not work for shared tiers, and as the DB was unresponsive I couldn’t run any queries to get active connections. This makes me vary about continuing on shared tier in the future, as the app is currently not yet in production but will be soon.

Is there anyone else who can share their experiences related to this?

Hi @Kormakur_N_A,

I would contact the Atlas in-app chat support regarding this. The team would have more insight into your project / cluster where the connection spike occurred. Please provide them with the cluster name / link when you open a chat with them and the same time frames when the connection spike occured similar to what you have provided here on the post.