Is there a way to turn off replication
Is there a way to turn off replication from one of the central stores, and only keep things in the EU? (GDPR compliance)
5 Replies
No, that's not possible currently. Can you tell us more about how you're planning to use KV? Lots of the value comes from the built-in caching for global distribution so I'm curious as to how you're using it/if there's a better alternative
Basically a user is able to upload files to a collection, and create a short link to that collection. I store the files in their region (R2), and store the link (KV) that has to reference their userId.
If I'm understanding GDPR correctly, this happens to be personally identifiable information since it could theoretically trace back to the EU jurisdiction bucket, and isn't compliant IF I am storing EU users files in a non EU jurisdiction DB.
Essentially it would work out fine if there was a way to disable the replication and create an EU only KV namespace -- but since that's not available, I was thinking about using an EU Jurisdiction R2 bucket & Worker Cache API as an EU only KV namespace.
Most of this is still hypothetical. I heavily use D1 for a majority of this project but I do not think it's appropriate for this since I may run into a 10GB D1 database limit (even though that would still get me plenty of rows)
Thanks. How would KV's built-in caching apply to your scenario? Are you looking to not have results cached in various data centers (in addition to the data not being stored in datacenters)?
From my understanding, caching is fine under legitimate use, as it is pull-based rather than push-based, and as long as the files/links are actually only ever stored in the EU for GDPR compliance. I do think I will end up using R2 for this
Got it, thanks @o7