CODEWITHSUNDEEP
google-app-engineauthenticationgoogle-cloud-platformgoogle-cloud-datastoregoogle-cloud-sql
snth
snth

Reputation: 3

Best storage option for user credentials (Google Cloud ecosystem)

In the Google Cloud ecosystem, what is the canonical storage option for basic user data including login credentials?

Let's optimistically say my app has 100K users, and I don't want to enable social login, but might want to in the future. I'm looking at CloudSQL and Datastore at the moment, and am unsure about the tradeoffs.

Upvotes: 0

Views: 141

Answers (1)

MrtN
MrtN

Reputation: 136

The main reason why I have been choosing Cloud Datastore in favour of Cloud SQL is the horizontal scalability of Datastore which is fully managed by Google. Since it is backed by Bigtable, you don't have to worry about scaling your instances.

Another thing you have to keep in mind is the cost model: with Cloud Datastore you pay per read/write/delete and storage. And for Cloud SQL you pay per running instance and storage. When you are starting up a website with only a few users, Cloud Datastore is very cheap, because you don't have to pay for a 24/7 running instance of Cloud SQL. When the traffic grows, it depends on the usage pattern what technology will be more cost-effective.

With Cloud SQL, you are more flexible when performing ad-hoc queries: you don't always have to add an index beforehand. And even if you do, you don't have to re-write all entities to add the entity (like it is the case in Cloud Datastore). And of course you have the obvious advantages: ACID support and a wide range of frameworks to choose from to interact with MySQL or PostgreSQL.

Keep in mind that both technologies seem similar, but the characteristics are completely different. You will encounter that when designing your Datastore dataset for scale and reading the documentation about the limits.

Finally, it might also be worth to take a look at these two products:

Upvotes: 1

Related Questions