[APM] Service inventory: detailed stats fetched for all services

[cauemarcondes]

closes #133697

Service Inventory

service.inventory.mov

Errors Overview

Before:

After:

[elasticmachine]

Pinging @elastic/apm-ui (Team:apm)

[cauemarcondes]

@elasticmachine merge upstream

[ogupte]

LGTM

[sorenlouv]

This might be technically correct but it's not super clear what this verifies. The test says "calls detailed API with visible items only".

There are 5 visible items afaict on line 140. Yet, detailed statistics requests 10 values. Isn't that a disconnect? Should those two be the same?

[cauemarcondes]

No, there are 5 pages, with 10 items on each of them.

[sorenlouv]

Okay I missed that. But that's also my point: it's not clear what this is verifying (at least not to me). Why are we asserting the number of pages in the pagination? Is that important when ensuring that it only makes requests for the visible items? Perhaps it's useful to assert that there are more pages than the current one.
Most importantly, I think you should make the connection between pageSize and groupIds more obvious:

    const pageSize = 10;
    cy.visit(`${javaServiceErrorsPageHref}&pageSize=${pageSize}`);
    cy.wait('@errorsMainStatistics');
    cy.get('.euiPagination__list').children().should('be.gt', 1);
    cy.wait('@errorsDetailedStatistics').then((payload) => {
      expect(payload.request.body.groupIds.length).eql(pageSize)
      );
    });

[sorenlouv]

Same as above. I'd expect this to be something like:

Suggested change

	expect(payload.request.body.groupIds).eql(
	JSON.stringify([
	'000000000000000000000000Error 10',
	'000000000000000000000000Error 11',
	'000000000000000000000000Error 12',
	'000000000000000000000000Error 13',
	'000000000000000000000000Error 14',
	'000000000000000000000000Error 15',
	'000000000000000000000000Error 16',
	'000000000000000000000000Error 17',
	'000000000000000000000000Error 18',
	'000000000000000000000000Error 19',
	])
	expect(payload.request.body.groupIds.length).eql(NUMBER_OF_VISIBLE_ERROR_GROUPS)

[cauemarcondes]

check the comment above.

[sorenlouv]

It looks like items only contains the current page's group id's. That's fine but different from how this is handled in error_group_overview/index.tsx. Can we handle this more consistently?

[cauemarcondes]

That's why error_group_overview uses ManagedTable and service_overview_errors_table uses EuiBasicTable. I personally don't like the ManagedTable since it introduces a bug if it is used more than one time on the same page since it updates the table options in the URL (pageIndex, pageSize, sortField and sortDirection).

[cauemarcondes]

There's an open issue to delete the ManagedTable #111328

[kibana-ci]

💚 Build Succeeded

• Buildkite Build
• Commit: 750a980
• Storybooks Preview
• Webpack Bundle Analyzer

Metrics [docs]

Async chunks

Total size of all lazy-loaded chunks that will be downloaded as the user navigates the app

id	before	after	diff
apm	2.9MB	2.9MB	+951.0B

Unknown metric groups

ESLint disabled line counts

id	before	after	diff
apm	86	87	+1

Total ESLint disabled count

id	before	after	diff
apm	100	101	+1

History

• 💛 Build #52914 was flaky 386f892
• 💚 Build #52223 succeeded b249135
• 💔 Build #52201 failed c22296f
• 💔 Build #52187 failed 195c880
• 💚 Build #52156 succeeded c956d30

To update your PR or re-run it, just comment with:
@elasticmachine merge upstream