Example cost calculation for storing archive data in Azure Storage
Posted on: December 4, 2023I had to estimate the cost of storing archive data in Azure Storage for one very specific scenario. First I tried to estimate the cost using the Azure Storage pricing page, but it was quite hard to estimate the exact cost since I was not completely sure about the API call usage in the scenario.
I decided to test the scenario and see how much it costs. I used 1000 GiB as a test so that I could then quite accurately estimate the total cost for hundreds of terabytes. Remember that the cost is highly dependent on the scenario, so you should test your own scenario and see costs from your scenario. I documented all the steps in detail, so you can use that as a reference for your own cost calculation scenarios.
My tests are documented here:
JanneMattila/azure-storage-demos
If I would like to highlight one thing from the tests, it would be that you should
quite carefully plan the size of the archived files and how frequently you
need to fetch them after they have been archived.
That might mean that you package smaller files into larger files so that you
can operate on a smaller number of files. You can use as simple as zip to package those files.
Example: You generate daily hundreds of thousands of small files that you need to archive. It might make sense to package them to full daily archives before storing them into storage account. If you need to fetch those to investigate specific day, you can then fetch the whole archive in one go.
Here is an example of cost for 1000 GiB file rehydration from Archive to Hot tier by copying it to separate storage container (this is covered in the GitHub in more detail):
That of course impacts your archival strategy, if you need to rehydrate the data frequently and that suddenly comes your cost factor.
As always, understand your requirements, data usage patterns, data volumes, and test your scenarios to see how much it costs.
I hope you find this useful!