feat: Use WAL Manager

[grobinson-grafana]

What this PR does / why we need it:

This commit changes the RF-1 ingester to use the WAL Manager instead of flushCtx.

Which issue(s) this PR fixes:
Fixes #

Special notes for your reviewer:

Checklist

• Reviewed the CONTRIBUTING.md guide (required)
• Documentation added
• Tests updated
• Title matches the required conventional commits format, see here
  • Note that Promtail is considered to be feature complete, and future development for logs collection will be in Grafana Alloy. As such, feat PRs are unlikely to be accepted unless a case can be made for the feature actually being a bug fix to existing behavior.
• Changes that require user attention or interaction to upgrade are documented in docs/sources/setup/upgrade/_index.md
• For Helm chart changes bump the Helm chart version in production/helm/loki/Chart.yaml and update production/helm/loki/CHANGELOG.md and production/helm/loki/README.md. Example PR
• If the change is deprecating or removing a configuration option, update the deprecated-config.yaml and deleted-config.yaml files respectively in the tools/deprecated-config-checker directory. Example PR

[benclive]

Nice, this looks great! I left a few comments but nothing major so I'm happy to approve.

[benclive]

It might be more performant to use a pointer for Append here - unsafe.Sizeof says the AppendRequest is 80B and since it's passed by value it gets copied into the function call.
I'm not sure if this actually leads to any performance benefit in this case but might be worth a benchmark since it's called so often?

[grobinson-grafana]

A lot of the time &wal.AppendRequest{} will be turned into wal.AppendRequest{} at compile-time as it's faster to do this copy on the stack instead of allocate it on the heap. It's easier to show with an isolated example.

[grobinson-grafana]

Benchmark for using the stack:

go test -gcflags=-N -bench=. bench_test.go
goos: darwin
goarch: arm64
BenchmarkDoRequest-8   	573065010	         2.098 ns/op
PASS
ok  	command-line-arguments	2.663s

type AppendRequest struct {
	TenantID  string
	Labels    map[string]string
	LabelsStr string
}

func doRequest(r AppendRequest) {
	_ = r
}

func BenchmarkDoRequest(t *testing.B) {
	labels := map[string]string{"foo": "bar"}
	labelsStr := "{foo=\"bar\"}"
	for i := 0; i < t.N; i++ {
		doRequest(AppendRequest{
			TenantID:  "1",
			Labels:    labels,
			LabelsStr: labelsStr,
		})
	}
}

Benchmark for using the heap:

go test -gcflags=-N -bench=. bench_test.go
goos: darwin
goarch: arm64
BenchmarkDoRequest-8   	346227085	         3.131 ns/op
PASS
ok  	command-line-arguments	1.654s

type AppendRequest struct {
	TenantID  string
	Labels    map[string]string
	LabelsStr string
}

func doRequest(r *AppendRequest) {
	_ = r
}

func BenchmarkDoRequest(t *testing.B) {
	labels := map[string]string{"foo": "bar"}
	labelsStr := "{foo=\"bar\"}"
	for i := 0; i < t.N; i++ {
		doRequest(&AppendRequest{
			TenantID:  "1",
			Labels:    labels,
			LabelsStr: labelsStr,
		})
	}
}

[benclive]

I think we could remove the err return from NextPending if we don't look at it?

[grobinson-grafana]

I think so too, I'll do it for both NextPending and Put in another PR. 👍