Upgrade uses of configgrpc/confighttp

[jade-guiton-dd]

Description

Follow up to open-telemetry/opentelemetry-collector#14190, which introduced an API breaking change: instead of passing a host component.Host to configgrpc and confighttp's ToClient / ToServer methods, components should instead pass the result of host.GetComponents(). Tests that don't make use of authentication or middleware extensions, as well as custom component hosts which don't support extensions (such as the OpAMP supervisor), can simply pass nil instead of using a no-op host.

Replacing host by host.GetComponents() is fairly straightforward, as long as you don't pass in a nil host, which unfortunately a lot of tests do, so I had to fix that as well.

This will be merged into the update-otel PR at #44581.

[jade-guiton-dd]

Pinging Sumologic extension code owners: @rnishtala-sumo @chan-tim-sumo @amdprophet

The TestUpdateMetadataRequestPayload test in extension/sumologicextension/extension_test.go calls SumologicExtension.getHTTPClient without calling SumologicExtension.Start first, which leads to confighttp.ClientConfig.ToClient being called with a nil host. This used to work as long as no authentication/middleware was configured, but it poses a problem now that components are expected to call host.GetExtensions() upfront, before calling ToClient; the most straightforward fix of the extension code leads to a segfault in the test.

I tried adding a call to Start in between, but I think the heartbeat loop started in the process causes the test to fail very verbosely. So I just manually set SumologicExtension.host to a no-op host instead. This is not a very elegant fix, so I will leave it to y'all to figure out what you want to do long term. Presumably, this would involve either:

• considering a nil SumologicExtension.host to be a valid state, and putting a condition around the call to GetExtensions(), or
• modifying the test to start the component with a valid host, but in a way which doesn't fail the test.

[jade-guiton-dd]

After the last commit, the scoped tests failed twice, but because of flaky tests that seem unrelated to the changes, so I'm marking this ready for review.

I have no idea why changing nil to componenttest.NewNopHost() caused a failure in processor/metricsgenerationprocessor/processor_test.go, considering the component doesn't seem to define a Start handler or use the host at all... It was an unnecessary change for the purposes of this PR so I reverted it, but maybe something to investigate.

[mx-psi]

I am going to merge before scoped tests pass since this is a PR against another PR