Dataverse Storage calculation

[JAWL]

Dear All, in the spirit of "there is no stupid question", anyone can provide me some insights how many rows (assume 10 columns and no media attachment) in Dataverse will make up 1 GB storage? FYI, I need to understand the potential cost for cost center manager to review before creating Dataverse Custom Environment.

[ChrisPiasecki]

Hi @JAWL,

It's difficult to estimate, as it will depend on the column data types. My suggestion would be to spin up a Sandbox environment, create the table and columns with the data types needed. You can then use a tool like Mockaroo Data Munger for Xrmtoolbox to generate a bunch of sample data. Then you can view capacity used by the Dataverse table in the Power Platform Admin Center and get your estimate there. Keep in mind there will be some capacity used by the metadata for all the tables and other components in the database so make sure to factor that into your estimate.

---
Please click Accept as Solution if my post answered your question. This will help others find solutions to similar questions. If you like my post and/or find it helpful, please consider giving it a Thumbs Up.

[rampprakash]

Hello @JAWL,

Adding to @ChrisPiasecki, there is post written by @dpoggemann for Storage for his one of the customer

https://powerusers.microsoft.com/t5/Microsoft-Dataverse/Dataverse-Database-capacity-and-file-capacit...

Note:

if you want to see Data overview then we need to go to 

1. https://admin.powerplatform.microsoft.com

2. Expand Resources then Capacity 

there you can see the capacity of Database based on the Entity

Please mark as Answer if it is helpful and provide Kudos

Subscribe : https://www.youtube.com/channel/UCnGNN3hdlKBOr6PXotskNLA
Blog : https://microsoftcrmtechie.blogspot.com

You may have 10 Fields in a Entity or 15/20/30 that depends and any one contains value and many cannot, so we cannot predict how much a size will be approximately the

[JAWL]

@rampprakash @ChrisPiasecki  Unfortunately, the Dataverse environment is controlled by a central team in my company. Thus, I'm not Admin and not authorized to access Admin Center. Is that the only way? Alternatively, I take the liberty to suggest if anyone of you can share some brief insights based on your Dataverse tables that you have created or are currently working on? 

[dpoggemann]

Hi @JAWL ,

This is really hard to approximate because it is based on the types of columns in the tables and this will vary table by table of course.  I calculated the size per custom column on a number of tables based on the size of the tables in MB, number of records, and number of columns and came out with the following (per column):

.00008938775698

.000036153644811

.000077673488898

.0000208505367465

Overall I think a safe guess would be something like .000075 for each custom column in your table so you could calculate with this formula:  .000075 per column * 10 columns * X rows = 1000 which would be around 1,333,333 records.  The challenge is the range, even based on above could be around 1.1 million records to 3.5 million records.

Hope this helps approximate in some way but not specifically a science at all...

[dpoggemann · Accepted solution]

Hi @JAWL ,

This is really hard to approximate because it is based on the types of columns in the tables and this will vary table by table of course.  I calculated the size per custom column on a number of tables based on the size of the tables in MB, number of records, and number of columns and came out with the following (per column):

.00008938775698

.000036153644811

.000077673488898

.0000208505367465

Overall I think a safe guess would be something like .000075 for each custom column in your table so you could calculate with this formula:  .000075 per column * 10 columns * X rows = 1000 which would be around 1,333,333 records.  The challenge is the range, even based on above could be around 1.1 million records to 3.5 million records.

Hope this helps approximate in some way but not specifically a science at all...

[JAWL]

@dpoggemann Appreciate your detailed answers! Let me take the extreme bottom (1 GB storage for 1million Dataverse records) or even if it's 1 GB storage for 500k will be approved by my cost center manager (because our monthly transaction is ~3k to 5k new records).

[ChrisPiasecki]

Hi @JAWL,

You don't need to be a tenant wide or service admin to access the admin center. You'll just be restricted to any admin functions or analytics for the Environments you're an admin on. You could spin up a developer or trial environment, or worst case a trial/dev M365 tenant and do your tests there.

Taking some arbitrary average column size off a dataset that looks nothing like your target will not give you the insight needed. A row with 10 boolean columns vs a row with 10 single line of text with 100+ characters could be an order of magnitude different, especially for tens or hundreds of thousands of rows.

---
Please click Accept as Solution if my post answered your question. This will help others find solutions to similar questions. If you like my post and/or find it helpful, please consider giving it a Thumbs Up.

[VNJoe]

The answer is "When you create your first environment, it automatically consumes the 'free' 1GB of space" and it ticks on logs as well so you'll be getting nagging emails the first weekend after you start using it.

Please mark this as answer to prove the point that 1 GB provisioned inefficiently is not nearly enough to even begin testing this product. They need to do a better job here. Thanks!