@@ -8,7 +8,9 @@ weight = 30
88
99# /ARCHITECTURE
1010
11- ### /DAG
11+ ---
12+
13+ ### DAG = Directed Acyclic Graph
1214
1315``` mermaid
1416graph TD
@@ -29,10 +31,7 @@ graph TD
2931 F --> G
3032```
3133
32- ---
33-
34- - DAG = Directed Acyclic Graph
35- - Graph to model dependencies and workflows
34+ <!-- - Graph to model dependencies and workflows -->
3635- Directed: The connections (edges) between steps (nodes) point in one direction — from one step to another.
3736- Acyclic: No cycles — you can’t return to a previous step by following edges.
3837
@@ -49,7 +48,7 @@ graph TD
4948### /DAGGER ENGINE
5049
5150- The Dagger engine is a custom version of BuildKit
52- - It is responsible for efficiently running your pipeline as a DAG (Directed Acyclic Graph)
51+ - It is responsible for efficiently running your pipeline as a DAG
5352- It's shipped as a container image and runs as a privileged container.
5453
5554---
@@ -79,37 +78,17 @@ The Dagger CLI is the interface between you and the Dagger engine. It's used to
7978
8079---
8180
82-
83- ### /Sample Dagger DAG (Mermaid)
84-
85- ```
86- graph TD
87- A[Checkout Code] --> B[Build Container]
88- B --> C[Test Application]
89- C --> D1[Run Linter]
90- C --> D2[Publish to Registry]
91- ```
92-
93- - Cloning code from a repo
94- - Building a container image
95- - Running tests
96- - Running a linter and publishing the image in parallel after tests
97-
98- ---
99-
10081### /🎯 Why GraphQL?
101- GraphQL is strongly typed and introspectable (perfect for generating SDKs).
102-
103- Lets you dynamically compose and query complex objects (like containers and filesystems).
104-
105- Supports lazy evaluation — only the final outputs you need get computed.
82+ - You write code in Go/Python/Node using a Dagger SDK.
83+ - That code sends GraphQL queries to a running Dagger engine (inside a container).
84+ - The engine interprets the queries, builds the DAG of container operations, executes it, and returns results.
10685
10786---
10887
10988### /🎯 Why GraphQL?
110- - You write code in Go/Python/Node using a Dagger SDK .
111- - That code sends GraphQL queries to a running Dagger engine (inside a container ).
112- - The engine interprets the queries, builds the DAG of container operations, executes it, and returns results .
89+ - GraphQL is strongly typed and introspectable (perfect for generating SDKs) .
90+ - Lets you dynamically compose and query complex objects (like containers and filesystems ).
91+ - Supports lazy evaluation — only the final outputs you need get computed .
11392
11493---
11594
@@ -145,13 +124,6 @@ Streams the output back to your SDK
145124
146125
147126
148- ### 🔄 Summary
149127
150- | **Concept** | **Dagger** |
151- |------------------|----------------------------------------------------------------------------|
152- | **DAG** | Internal representation of all operations and dependencies |
153- | **GraphQL API** | How your local code communicates with the Dagger engine |
154- | **SDK** | Thin client that generates and sends GraphQL queries |
155- | **Engine** | Containerized GraphQL server that builds, runs, and caches workflows |
156128
157129{{% /section %}}
0 commit comments