Home›Google Cloud in 5 Days›Day 2
Day 2 of 5
⏱ ~60 minutes
Google Cloud in 5 Days — Day 2

Compute: Cloud Run and Compute Engine

Deploy a containerized app to Cloud Run (serverless) and a traditional VM to Compute Engine. Understand when each is the right choice.

Cloud Run: Serverless Containers

Cloud Run runs Docker containers without managing servers. It scales to zero (no traffic = no cost) and handles scaling automatically.

Dockerfile
FROM python:3.11-slim
WORKDIR /app
COPY requirements.txt .
RUN pip install -r requirements.txt
COPY . .
CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "8080"]
Deploy to Cloud Run
# Build and push to Artifact Registry
gcloud run deploy my-service \
  --source . \
  --platform managed \
  --region us-central1 \
  --allow-unauthenticated

# Or build/push manually
gcloud builds submit --tag gcr.io/PROJECT_ID/my-app
gcloud run deploy my-app --image gcr.io/PROJECT_ID/my-app
Compute Engine VM
# Create a VM
gcloud compute instances create my-vm \
  --machine-type e2-micro \
  --zone us-central1-a \
  --image-family debian-12 \
  --image-project debian-cloud

# SSH into it
gcloud compute ssh my-vm --zone us-central1-a

# Stop when done to avoid charges
gcloud compute instances stop my-vm --zone us-central1-a
ℹ️
Cloud Run is the right choice for stateless HTTP services — APIs, web apps, microservices. Use Compute Engine when you need persistent disk, specific OS configuration, or long-running non-HTTP workloads.
📝 Day 2 Exercise
Deploy an App to Cloud Run
  1. B
  2. u
  3. i
  4. l
  5. d
  7. a
  9. s
  10. i
  11. m
  12. p
  13. l
  14. e
  16. P
  17. y
  18. t
  19. h
  20. o
  21. n
  23. F
  24. l
  25. a
  26. s
  27. k
  29. o
  30. r
  32. N
  33. o
  34. d
  35. e
  37. E
  38. x
  39. p
  40. r
  41. e
  42. s
  43. s
  45. a
  46. p
  47. p
  48. .
  50. C
  51. o
  52. n
  53. t
  54. a
  55. i
  56. n
  57. e
  58. r
  59. i
  60. z
  61. e
  63. i
  64. t
  66. w
  67. i
  68. t
  69. h
  71. D
  72. o
  73. c
  74. k
  75. e
  76. r
  77. .
  79. D
  80. e
  81. p
  82. l
  83. o
  84. y
  86. t
  87. o
  89. C
  90. l
  91. o
  92. u
  93. d
  95. R
  96. u
  97. n
  98. .
  100. V
  101. e
  102. r
  103. i
  104. f
  105. y
  107. t
  108. h
  109. e
  111. p
  112. u
  113. b
  114. l
  115. i
  116. c
  118. U
  119. R
  120. L
  122. w
  123. o
  124. r
  125. k
  126. s
  127. .
  129. C
  130. h
  131. e
  132. c
  133. k
  135. t
  136. h
  137. e
  139. l
  140. o
  141. g
  142. s
  144. i
  145. n
  147. t
  148. h
  149. e
  151. c
  152. o
  153. n
  154. s
  155. o
  156. l
  157. e
  158. .

Day 2 Summary

  • Cloud Run: serverless, scales to zero, charges only for requests. Perfect for APIs.
  • Compute Engine: full VMs, persistent disk, full OS control. More expensive, more flexible.
  • gcloud run deploy --source . auto-builds and deploys from source code.
  • Always stop VMs when not in use. Compute Engine charges by the second even when idle.
→
← Previous
Day 1: GCP Console and Core Concepts
Next →
Day 3: Storage: GCS and Cloud SQL
→
Finished this lesson?