How to Set Concurrency in Cloud Run for Google Cloud

To set concurrency in Cloud Run, use the --concurrency flag with the gcloud run deploy command or set it in the Cloud Console. This controls how many requests each container instance can handle at the same time, improving resource use and performance.

Syntax

The concurrency setting controls how many requests a single Cloud Run container instance can process simultaneously.

Use the following syntax with the Google Cloud CLI:

• gcloud run deploy SERVICE_NAME --concurrency=NUMBER

Where:

• SERVICE_NAME is your Cloud Run service name.
• NUMBER is the maximum concurrent requests per container instance (1 to 1000).

bash

gcloud run deploy SERVICE_NAME --concurrency=NUMBER

Example

This example deploys a Cloud Run service named hello-world with concurrency set to 10. This means each container instance will handle up to 10 requests at the same time.

bash

gcloud run deploy hello-world --image=gcr.io/cloudrun/hello --concurrency=10 --region=us-central1 --platform=managed

Output

Deploying container to Cloud Run service [hello-world] in project [YOUR_PROJECT_ID] region [us-central1]
Done.
Service [hello-world] revision [hello-world-00001-abc] has been deployed and is serving 100 percent of traffic.

Common Pitfalls

Common mistakes when setting concurrency include:

• Setting concurrency too high can cause slow response times if your container cannot handle many requests at once.
• Setting concurrency to 1 disables request multiplexing, which can increase costs.
• Forgetting to specify the --concurrency flag defaults concurrency to 80.
• Not matching concurrency settings with your container's actual ability to handle parallel requests.

bash

Wrong:
gcloud run deploy my-service --concurrency=0

Right:
gcloud run deploy my-service --concurrency=1

Quick Reference

Option	Description	Notes
--concurrency=NUMBER	Sets max concurrent requests per container instance	Valid range: 1 to 1000
Default	80	If not set, concurrency defaults to 80
Concurrency=1	One request at a time per instance	Good for non-thread-safe apps
Concurrency>1	Multiple requests at once	Better resource use if app supports it

Key Takeaways

Use the --concurrency flag with gcloud run deploy to set concurrency.

Concurrency controls how many requests a container instance handles at once.

Set concurrency based on your app's ability to handle parallel requests.

Default concurrency is 80 if not specified.

Concurrency=1 means one request at a time, which can increase costs.