GCP Compute, Storage and BigQuery
Three GCP services that map to AWS EC2, S3 and Redshift — but often with a friendlier developer experience.
GCP Core Services
Compute Engine — VMs
Same idea as AWS EC2:
gcloud compute instances create web-01 \
--zone=asia-south1-a --machine-type=e2-micro --image-family=debian-12
Special sauce: live migration — Google can move a running VM between hosts without downtime. AWS cannot do that.
Cloud Storage — object storage
Buckets + objects, same idea as S3. Storage classes:
- Standard (hot data),
- Nearline (accessed monthly),
- Coldline (accessed yearly),
- Archive (legal retention).
gsutil mb -l asia-south1 gs://my-notes-bucket
gsutil cp lecture.pdf gs://my-notes-bucket/
BigQuery — SQL on petabytes
BigQuery is a serverless data warehouse. Load billions of rows, query with plain SQL, pay per TB scanned.
SELECT category, COUNT(*) AS n
FROM `project.dataset.articles`
GROUP BY category
ORDER BY n DESC
LIMIT 10;
Real magic: no server to provision, no cluster to size. You literally just SQL.
Real-world example
A college data-science project: dump a 500GB CSV into Cloud Storage, load into BigQuery in 30 seconds, run all your analysis in the browser. Cost for the whole semester: under $5.
Keep reading
You may also like
gcp
Cloud Run vs GKE: When to Choose What
A pragmatic decision framework for serverless containers on Google Cloud.
gcp
Google Cloud for Beginners: Projects, IAM and Cloud Shell
How GCP organizes accounts, why 'projects' are important, and the free browser terminal that has everything pre-installed.
aws
ECS Part 3: Deploy to Fargate — Cluster, Task, Service
Create the cluster, define the task, run it as a service, open the security group — your container is live.
Discussion (0)
No comments yet. Be the first to weigh in.