How to get estimation for my org's requirements

I need details about pricing on managed mongo db instance on azure. I have gone through the Azure pricing sheet and have some questions. We need the estimated resources that will be required to store at least a million records and sustain certain number of requests per minute. Is the pricing based per database/collection/throughput ?. We also want to know what services would be included in the managed instance.

We have tried reaching out the support team to help us out in choosing the correct tier of Dedicated+Managed clusters but no response in 4 hours. I have seen the pricing sheet and want to know which tier would be suitable for us.

Hi @Vishal_Mishra,

The Atlas tier is based on the managed instance type similar to VM tiering. The instance types differ per CPU, Storage, RAM and amount of connection limits/network capabilities.
Cluster price can also differ by regions and by amount of nodes you run.

The following docs should help you started

I suggest you keep this effort with support to better fit it for your needs


Also, would it be possible to talk to a representative regarding the best options to purchase the required tier; our need is somewhat urgent

Hi @Pavel_Duchovny ,
Thanks for the response. There are a few things that are still unclear. Say I have a requirement to store 50 million records with 100K incoming requests per minute, what tier should I choose for Azure ? Should the M60 tier suffice in this case or should I opt for M80 tier ? I understand the cluster and storage auto scaling part, but I want to start with a price estimation as I saw on pricing sheet.


Hi @Vishal_Mishra,

The needed size is a subject of load testing for your secific use case. The main points is storage should be sufficient to store the 50M documents and Memory big enough to keep your “hot” working set. However, for one document type it can be 50GB but for another it can be 500GB or 1TB so its hard to guess.

One of Atlas advantages is you can start your storage and compute tier at one place and transparently, if application considerations are respected, scale your cluster to fit the growth requirements as you go.

We have a great serious of blogs for schema design and performance best practices which I recommend:

Since sizing your production is a complex task we offer consultancy packages for this.

Best regards,