AWS CodeBuild is slow in ap-southeast-1
Hello AWS Community,
I'm facing an issue with AWS CodeBuild's provisioning phase in ap-southeast-1 (public projects). My builds, used for Terraform deployments, are now taking 4-5 minutes for provisioning alone, whereas they used to complete in less than a minute. This slowdown is recent, despite no changes to my CodeBuild configuration.
What could be causing this increased provisioning time, and are there any strategies or best practices to accelerate CodeBuild provisioning in this region?
- Neueste
- Die meisten Stimmen
- Die meisten Kommentare
There are several potential reasons for the increased provisioning time in your AWS CodeBuild project in ap-southeast-1:
-
Regional capacity or service issues: There might be temporary service issues or capacity constraints in the ap-southeast-1 region affecting CodeBuild provisioning times.
-
Concurrent build limits: If you're running multiple builds simultaneously, you might be hitting concurrent build limits, causing builds to queue and wait for resources.
-
Container image pull delays: If your build is using a container image that needs to be pulled each time, this could be causing delays in the provisioning phase.
Here are some strategies to improve provisioning times:
-
Implement caching: Use CodeBuild's caching features to cache Docker layers and dependencies. You can configure local caching with modes like LOCAL_DOCKER_LAYER_CACHE to avoid pulling images repeatedly.
-
Increase computational power: Consider using a more powerful compute type for your build environment to speed up the overall build process.
-
Check concurrent build settings: Verify your concurrent build limits and increase them if necessary to prevent builds from waiting in a queue.
-
Run builds in parallel: Take advantage of CodeBuild's ability to run multiple builds in parallel rather than sequentially.
-
Consider region proximity: If possible, use a region that's geographically closer to your location or to where your resources are deployed, as network latency between distant regions can impact performance.
If the issue persists and seems to be a service-side problem, you might want to check the AWS Health Dashboard for any ongoing issues with CodeBuild in your region.
Sources
Pipeline build taking more time for ECS using codepipeline | AWS re:Post
CodeBuild Machine is Failing in Provisioning phase | AWS re:Post
Build phase - Development and Test on Amazon Web Services
Troubleshooting AWS CodeBuild - AWS CodeBuild
Pull and image and cache in codebuild | AWS re:Post
Also experienced this in us-east-1, Issue is now resolved for me in us-east-1.
Are you still facing this in ap-southeast-1?
