Fauna Data Manager Pricing

Hi all, I’m new to Fauna and am researching if it would be a good fit for a new project I am starting.

I have a need to provide exports of entire collections for use as backups periodically. Rather than loop through all of the records in a pagination/get query, it seems like I could just use FDM to export the entire collection to a JSON file.

My question is if doing that is the best practice, and also how it would be impacted by billing? I don’t want to incur millions of reads every time I want to export all of the records.

Similarly, I would need the ability to import large collections from time to time (again, possibly millions of documents). If I use FDM, would I be charged for a write for each document?

I couldn’t find anywhere in the documentation that outline how FDM imports and exports are billed in terms of reads and writes.

4 Likes

Everything is (currently) billed as if you were a client, because under the hood, FDM is just using the client interface. How often will you be backing up? Currently backing up a million docs will cost you $0.50.

Thanks for your answer Ben. So it sounds like FDM actually just uses reads/writes for each document being imported/exported, correct? In my case, I am moving from an existing system with a lot of data over to Fauna.

So, for example, if I have 250 million records to import, that would cost about $500 for the initial import?

And then if I am doing daily backups, would that be a read operation for each record as well? That seems like it would cost about $125 a day just for backups.

This looks accurate to me. You could probably chop down your dailies significantly by using some kind of temporal index. I think for projects with these kind of volumes we need dedicated tooling. I see you’ve been chatting with @summer on slack, so this will bubble back up to the right people I’m sure.

Thanks for confirming Ben, and yeah I was thinking something similar for the backups approach, it’s kind of a paradigm shift from the MySQL world but that’s not necessarily a bad thing.

I’d be very interested in seeing if something could be added in the future to handle full disk snapshots or something similar for backups to make those cheaper, and if we could cut down the cost of initial imports somehow that could encourage more people to make the switch by lowering the barrier to getting data in and using fauna, which could be a win-win.

1 Like

@Leighton_Whiting I’m facing the same situation. I’ve posted a similar questions:

Export big collections via CDC or dump to S3 or Snowflake

Did make any advances on the subject or had any revelations :slightly_smiling_face:?

It seems like FDM uses regular read/write quotas and pricing. I ended up going with a different provider unfortunately. FaunaDB seems really awesome, and I’ll definitely use it for other projects in the future, but for this project it didn’t make sense due to these limitations.

1 Like