GCP Databases in a Nutshell
#Google Cloud
#Professional Data Engineer Certfication
#Database
#Cloud SQL
#Bigtable
#Spanner
#Firestore
Intro
This post is part of a series of posts with notes as I’m studying for Google’s Professional Data Engineer Certification.
This particular post covers Cloud Composer in a nutshell.
Disclaimer
Please read this disclaimer.
Databases
Grouped by Management Model
Managed Services
- Cloud SQL
- Cloud Spanner
- Cloud Bigtable
Serverless Services
- Cloud Storage
- Firestore
Grouped by Size/Relationality
| Size of Data: | Small | Large |
|---|---|---|
| Relational | Cloud SQL | Spanner* |
| Non-Relational | Firestore** | Bigtable |
* The option of using processing units as the means of calculating cost for Spanner makes it a better option for smaller solutions ** Firestore can handle large amounts of data too when provisioned correctly, but it can be used for small applications where-as Bigtable really isn’t/shouldn’t be used for small use-cases
Grouped by Proprietary-ness
Google Proprietary
- Spanner
- Firestore
Generic
(data in these systems could be easily transfered on-prem. or to a diff. cloud provider)
- SQL
- Bigtable (h-base compatible)
Transaction Support
ACID Transactions
- SQL
- Spanner
- Firestore
Row-Level Transactions
- Bigtable