Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix multi-module jar skip for spring-boot projects #1810

Merged
merged 1 commit into from
Apr 19, 2024

Conversation

awisniew90
Copy link
Contributor

No description provided.

Copy link
Member

@cherylking cherylking left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Just the one question.

Copy link
Member

@scottkurz scottkurz left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think it's worth adding an IT here. We have the whole structure in src/it/dev-it/src/test/java/net/wasdev/wlp/test/dev/it... and the validation can be pretty simple to simply check that it wasn't skipped.

@awisniew90
Copy link
Contributor Author

I think it's worth adding an IT here. We have the whole structure in src/it/dev-it/src/test/java/net/wasdev/wlp/test/dev/it... and the validation can be pretty simple to simply check that it wasn't skipped.

So basically run "liberty:dev" against a spring boot app. I only see two existing spring boot apps in the test source here, and none in the dev it structure. So we'd need to create a Spring Boot app in dev-it and validate that we are not skipping the execution.

@awisniew90 awisniew90 force-pushed the fix-sb-multi branch 2 times, most recently from 33453ac to 5c45421 Compare April 15, 2024 17:54
public class SpringBootDevTest extends BaseDevTest {

@BeforeClass
public static void setUpBeforeClass() throws Exception {
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You have to do more like mvn package liberty:dev.

That said.. I think it might be a bad idea to include this test.. since someone later might assume it implies that we have support for dev mode in Spring Boot which we don't.

This would fit better as a test of 'run'. You still have to do a package with 'run' (so mvn package liberty:run) but you don't have to use loose apps, and so it makes more sense than dev (since neither loose apps in general nor dev work with SB and LMP currently).

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

^^ What Scott said :-)

@@ -1342,7 +1342,11 @@ private void doDevMode() throws MojoExecutionException {

// In a multi-module build, dev mode will only be run on one project (the farthest downstream) and compile will
// be run on any relative upstream projects. If this current project in the Maven Reactor is not one of those projects, skip it.
List<MavenProject> relevantProjects = getRelevantMultiModuleProjects(graph);
boolean skipJars = true;
if("spring-boot-project".equals(getDeployPackages())) {
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

At this point, is this just an error, independently of whether we think we should skip? Or are we trying to make this correct imagining a future effort to add support for Spring Boot in dev mode?

@@ -119,7 +119,11 @@ private void generateFeatures() throws MojoExecutionException, PluginExecutionEx

// In a multi-module build, generate-features will only be run on one project (the farthest downstream).
// If this current project in the Maven Reactor is not that project or any of its upstream projects, skip it.
List<MavenProject> relevantProjects = getRelevantMultiModuleProjects(graph);
boolean skipJars = true;
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Does "generate-features" do something meaningful today at LMP v3.10, in the Spring Boot JAR case? Are we adding any value at all by complicating the path here?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I doubt we have tested this scenario. So what would perform the most like it did before the changes went in to skip jar modules? Do we want to not skip jar modules at all for now for generate-features?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think that might cause problems since generate-features is called from dev mode... so in the non-springboot case, we could flag a conflict in generate-features (a jar and a war module without any downstream projects), that we didnt flag in the dev mode execution. So the two need to behave the same in that regard.

Since dev mode is not supported in the Springboot case, should we just add this SB check to the RunServerMojo and not for DevMojo and GenerateFeaturesMojo?

@scottkurz
Copy link
Member

scottkurz commented Apr 16, 2024

Sorry, not sure which comment to respond to so starting a new one:

  1. think it might be best to leave things alone and not add enforcement. In general it's nice to fail fast but a higher burden of proof that that'd be good now that it's out there.
    So while maybe I'd suggest failing fast for Spring Boot JAR in dev mode.. I'm backing off that.

  2. Another related change we could make it to skip what I'm calling the multi module "reducer" for a single-module case?
    So if this returns a one-item list (in getRelevantMultiModuleProjects)

List sortedReactorProjects = graph.getSortedProjects();

let's just return the single module and keep going. This would solve the original case, the SB guide..which is single module.
However, I think the fix Adam coded is better. One could have a JAR->JAR multi-module... and the original problem would remain. Might as well fix it, like the PR now does. Once that is fixed... is it worth also making the change here to short-circuit single module execution? I don't know. Could someone have done something with a JAR package that we aren't thinking of? Do we care about saving a few cycles? Eh..not sure it matters but helpful to have discussed here.

  1. If we liked the hierarchy better, we could move getDeployPackages() to the existing common base ServerFeatureSupport instead of moving GenerateFeaturesMojo in the hierarchy. I don't know the hierarchy looks better either way so no strong opinion.

@cherylking cherylking merged commit 9b96fb6 into OpenLiberty:main Apr 19, 2024
12 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants