Mongo Charts & Aggegation Pipelines timeouts

Hi there,

Any advice is highly appreciated. I’ve been running an v4.4 M20 Atlas Cluster happily for a few years now.

The Search, Charts, aggregation pipelines in Compass have all been working great - until recently.

The Charts & aggregation pipelines in Compass specifically have become so slow and subject to frequent timeouts.

I’ve tried creating Views to project data into the smallest format possible, but even sometimes View creation fails due to timeouts.

Memory pressure is not high on the cluster, just these general queries that used to be so performant have come to a crawl. Nearly all of my charts are broken.

I’m getting concerned that Mongo isn’t the right tool for scale, I need aggregations and charts to make informed decisions. Also, if the cluster is having a hard time performing with these queries, I’m worried as this database grows the performance will affect the production apps.

Any suggestions besides checking the performance advisor and creating indices?

Additional info:

  • Avg Doc Size: 37kb
  • ~700k records

May be an idea to supply example data along with queries you’re running as well as current indexes on the data if you want feedback on what could be causing an issue.

Have you tried running the queries with explain to check if indexes are being hit?

Assuming that indexes exist and are being used correctly then server logs from calls could help someone pinpoint issues with the configuration.

4.4 is also a very old version, it’s past EOL now:

complex aggregation pipeline and indexes can impact on the performance when dealing with the datasets. Simplify the query where possible and break them into small steps.