Cloudwithdj

Who owns AWS Cost optimization at your company?

AWS CodeBuild vs AWS CodePipeline, aren’t they both part of CI/CD?
Yes, they are. But they handle very different parts of the process.

Here’s how I explain it when someone asks, “Do I need both?”

✅ AWS CodeBuild
- Builds service: it compiles code, runs tests, creates artifacts
- Fully managed: you don't need to provision build servers
- Uses buildspec.yml to define build steps
- Supports Docker, Java, Python, Go, .NET, and more
- Scales automatically, pay-as-you-go
- Great for unit tests, packaging, static analysis

CodeBuild is the “do the work” part of CI.

✅ AWS CodePipeline
- Orchestration tool: it connects your stages (Source → Build → Deploy)
- Automates end-to-end delivery workflows
- Integrates with CodeCommit, S3, GitHub, CodeBuild, CodeDeploy, etc.
- Uses JSON or YAML definitions (via AWS Console or CLI)
- Each stage can use a different service (or custom actions)
- Monitors each step and stops on failure.

CodePipeline is the “glue” that manages the whole pipeline.

My take
- Use CodeBuild when you need to build and test your code.
- Use CodePipeline to connect everything together (source, build, deploy).

Most teams use both CodeBuild as a stage inside CodePipeline.

Are you using CodePipeline end-to-end, or just triggering builds/deploys manually?

S3 Glacier vs S3 Standard, which one should you choose?

This is how I break it down when folks ask me “Where should we store this data?”

✅ S3 Standard:
- Instant access, low latency
- Great for frequently accessed data
- Designed for real-time apps, websites, big data
- Lifecycle transitions make it easy to move to cheaper tiers later
- Higher cost, but you pay for speed

S3 Standard is ideal when availability and performance are top priorities.

✅ S3 Glacier
- Cold storage, long-term, infrequent access
- Retrieval can take minutes to hours (Glacier) or up to 12h (Deep Archive)
- Great for backups, archives, compliance, audit trails
- Much cheaper than Standard (cents on the dollar)
- Data retrieval charges apply, so plan access carefully

S3 Glacier is ideal for “store and forget” scenarios.

My opinion:
- Use S3 Standard when your app or users need that data now.
- Use S3 Glacier (or Deep Archive) when you won’t touch it for months or years.

Both are powerful.

But the right choice depends on access patterns, retention policy, and budget.

What’s your go-to S3 tier? How do you manage storage costs?

Read my newsletter for bi-weekly updates and new features from AWS, Iac, Cloud Natives and DevOps tools: whatsnewcloud.com

Visit my website: cloudwithdj.com