GCP Datastores

Question 1

What is MemoryStore for Redis?

Answer 1

Google Cloud’s fully managed Redis service.

Question 2

What is Cloud Storage?

Answer 2

Cloud Storage is a managed service for storing unstructured data as Objects/Blobs.

Question 3

What are the four storage options for Cloud Storage?

Answer 3

Standard - frequently accessed data

Nearline - infrequently accessed data (30d or less)

Coldline - infrequently accessed data (90d)

Archive - Rarely accessed data (1y)

Question 4

What is Object Lifecycle Management?

Answer 4

A Cloud Storage feature that allows you to configure how long a bucket keeps data in a certain storage option.

Example: Changing from Standard to Coldline after 2 weeks of inactivity.

Question 5

What is a retention policy?

Answer 5

A retention policy is a feature that specifics how long data must be unchanged for.

It is applied retroactively on existing objects in a bucket.

Question 6

What is a retention policy lock?

Answer 6

A guard that prevents a retention policy from being removed.

Once applied you cannot reduce or remove it, but you can increase it.

Furthermore, you cannot delete a locked bucket if it has objects that haven’t met the retention period.

Question 7

What is Redis?

Answer 7

An open-source, in-memory data store, typically used for caching data.

Question 8

What is BigQuery?

Answer 8

BigQuery is a serverless data warehouse with built-on ML/AI and BI.

Question 9

What is Object Versioning in Cloud Storage?

Answer 9

Object versioning is a Cloud Storage feature that tracks changes to objects within a bucket and allows admins to manage (restore, delete, etc) them.

Question 10

What is Firestore?

Answer 10

A serverless NoSQL database

Question 11

What is Cloud Spanner?

Answer 11

A managed globally available, SQL/relationship database with 99.999% availability.

Question 12

What is Cloud SQL?

Answer 12

A fully managed, regionally available SQL database with support for MySQL, PostgreSQL and Microsoft SQL.

Question 13

What is Filestore?

Answer 13

A fully managed shared filesystem from Google Cloud.

Question 14

What is Persistent Disk?

Answer 14

Block storage for VM instances. They act as the instance’s data drive.

Question 15

How are objects stored in Cloud Storage?

Answer 15

They are stored in buckets as immutable objects.

A ‘change’ to an object just creates a completely new one.

Question 16

What is the difference between Cloud SQL and Cloud Spanner?

Answer 16

Cloud Spanner is much more expensive as it offers high availability globally and massive amounts of data (PBs).

Cloud SQL is only available regionally and stores much less data (TB)

Question 17

What is Datastore?

Answer 17

The previous version of Firestore. It’s suitable for storing semi-structured data and has high consistency, but lacks Firestore’s real-time and offline capabilities.

Question 18

What is BigTable?

Answer 18

BigTable is a high performance, NoSQL database in the form of a sparsely populated table can scale up to millions of rows and thousands of columns.

It can store up to PBs of data.

Question 19

How can an application connect to a Cloud SQL instance?

Answer 19

Private IP
Cloud SQL Auth Proxy
Public IP (Cloud SQL Auth Proxy is required when using public IP)

Question 20

What is Cloud SQL Auth Proxy?

Answer 20

Cloud SQL Auth Proxy authorises and secures Cloud SQL connections by using IAM Permissions.

It validates IAM service accounts/users and secures the connection with a SSL/TLS layer. It is also the recommended way of connecting to a Cloud SQL Database because of the above.

Question 21

How you authorise to Cloud SQL?

Answer 21

Cloud SQL Language Connectors - Client libraries to connect a SQL DB

Cloud SQL Auth Proxy - Service to authorise and authenticate connections

Self-managed SSL/TLS certificates - User-issued certificates for data encryption

Authorized Networks - Add authorised IPs. (use with self-managed certs as a best practice)

Question 22

How much data does Cloud SQL store?

Answer 22

Upto 30TB of data.

Question 23

When should you use Cloud SQL?

Answer 23

When you want to:

• Migrate local MySQL, PostgreSQL, and SQL Server databases.
• To reduce maintenances costs for a pre-existing database

Question 24

When should you use Cloud Spanner?

Answer 24

When you want to:

• Store massive amounts of data
• Scale your database infinitely
• Support global expansion
• You need higher availability than Cloud SQL (three 9s vs .95)

Question 25

In Cloud SQL, when should you use Private IP connection over Public?

Answer 25

Private IP is the recommended choice as it’s the most secure (only clients on the same VPC can connect to it), but if the database consumers aren’t on a VPC, then use a public IP.

Question 26

In Cloud SQL, what do you need to configure in order to use Private IP?

Answer 26

Private Services Access as Cloud SQL cannot use Private Google Access.

Question 27

What is an ACL?

Answer 27

An Access Control List is a mechanism to define fine-grained access to Objects.
It is a legacy mechanism kept for interoperability with AWS, and uniform bucket-access (IAM) is recommended over it.

Question 28

What is Dataflow?

Answer 28

Dataflow is a fully-managed service for transforming and enriching data in stream (real-time) and batch modes. It provides a simplified pipeline development environment that uses the Apache Beam SDK to transform incoming data and then output the transformed data.

Source - https://cloud.google.com/dataflow/docs/tutorials/dataflow-stream-to-bigquery