Unterschiede

Hier werden die Unterschiede zwischen zwei Versionen angezeigt.

--- en:modul:m324_aws:topics:03 [2025/10/15 12:45] – angelegt dgaravaldi
+++ en:modul:m324_aws:topics:03 [2025/10/21 12:41] (aktuell) – [Setting agents (slave)] dgaravaldi
@@ Zeile 1: / Zeile 1: @@
 ===== Jenkins architecture =====
-\\
+<sup>Internal reference: topics/02-2.md</sup>
 \\
 ==== Master and agents ====
@@ Zeile 34: / Zeile 33: @@
 \\
 ==== Test and production instances ====
-How to test the Jenkins upgrades, new plugins, or pipeline definitions? Jenkins is critical to the whole company. It guarantees the quality of the software and, in the case of continuous delivery, deploys to the production servers. That is why it needs to be highly available, and it is definitely not for the purpose of testing. It means there should always be two instances of the same Jenkins infrastructure – test and production.
+How to test the Jenkins upgrades, new plugins, or pipeline definitions? Jenkins is critical to the whole company. It guarantees the quality of the software and, in the case of continuous delivery, deploys to the production servers. That is why it needs to be highly available, and it is definitely not for the purpose of testing. It means __there should always be two instances of the same Jenkins infrastructure__ – test and production.
 \\
@@ Zeile 51: / Zeile 50: @@
 Agents always communicate with the Jenkins master using one of the protocols: __SSH (sshd)__ or __Java web__ start. At a higher level, we can attach agents to the master in various ways:
-  * <strong>Static versus dynamic</strong>: The simplest option is to add agents permanently in the Jenkins master. The drawback of such a solution is that we always need to manually change something if we need more (or fewer) agent nodes. A better option is to dynamically provision agents as they are needed.
+  * **Static versus dynamic**: The simplest option is to add agents permanently in the Jenkins master. The drawback of such a solution is that we always need to manually change something if we need more (or fewer) agent nodes. A better option is to dynamically provision agents as they are needed.
-  * <strong>Specific versus general-purpose</strong>: Agents can be specific (for example, different agents for the projects based on Java 8 and Java 11) or general-purpose (an agent acts as a Docker host and a pipeline is built inside a Docker container).
+  * **Specific versus general-purpose**: Agents can be specific (for example, different agents for the projects based on Java 8 and Java 11) or general-purpose (an agent acts as a Docker host and a pipeline is built inside a Docker container).
 These differences resulted in four common strategies for how agents are configured: