M10 cluster slower than M5 with same database

Hi all,

We have an app on MongoDB, 15 collections with a few of docs, and one collection with about 5000 documents.

When we use a M10 dedicated cluster, it is so much slower than a M5 shared cluster with the exactly same database and indexes.

Our app is almost unusable on a M10 cluster.
We download and upload the database with mongodump and mongorestore tools.

I don’t see any issue on the M10 cluster dashboard, I did every indexes recommandations.

The M5 cluster has very good performances but has a lack of storage for our usage. So we realy need to upgrade to a more confortable plan.

What can be the reason of this slowness ?

Thanks

2 Likes

Interesting, did you find out why? (curios)

Is the disk size the same configured on the M5 vs M10? The IOPS are based off storage size. I’m assuming CPU is the same or better, and you’ve checked that.

This week I upgraded from m5 to m10 and I’m having the same problem, I started to notice slowness in my API routes, so I identified slow searches in the database, I worked for two days creating new indexes and comparison tests in my queries, but looking at the logs from a few days ago I noticed that the same queries in the same conditions are much, much slower, I don’t know what to do and I need to solve it. Have you had any success in finding the problem?