Insanely huge RPU because of indexes

Hi everyone,

I have an issue in which i had and extremely huge number of RPU operations when i created indexes.
I usually have 1K/s but today i reached over 150K/s which make me a lot worried for the cost of this. I deleted all indexes and things seems to be back to normal but i’m quite surprised because in guides that i read, they always say to use indexes to reduce cost. I want to know if another person fell into the same trap and what happened ? I’m very worried about my position in the company if we end up with huge cost because of this.
I also think that if i should be able to see my billing in real time and not waiting a day for it because being in this situation is quite uncomfortable.

Thanks for reading

I had those spikes again and i’m now very very worried. I deleted a time series collection that i created to see if it was this that created those huge rpu.

Hey Valentin,

Please reach out to our support team either via a support ticket or the chat icon in the bottom right corner of the Atlas UI. We’ll be sure to help.

Also feel free to DM me your serverless instance; I’m happy to take a look. Typically, indexes will reduce the number of RPUs consumed (see Serverless Instances Billing 101: How to Optimize Your Bill with Indexing | MongoDB), though we can certainly dig deeper.

Best,
Chris
Atlas Serverless Product Manager

1 Like