[chore] update jaeger dep and remove koanf 1.5 dep

[codeboten]

Fixes #28647

After this is merged contributors can finally use go workspaces in this repo.

Fixes #26567

[github-actions]

This PR was marked stale due to lack of activity. It will be closed in 14 days.

[codeboten]

not stale, waiting on PR

[github-actions]

This PR was marked stale due to lack of activity. It will be closed in 14 days.

[codeboten]

There's a change in jaeger v1.45.0 that causes the pkg/translator/jaeger test to fail:

--- FAIL: TestProtoToTraces (0.00s)
    --- FAIL: TestProtoToTraces/two-spans-with-follower (0.00s)
        jaegerproto_to_traces_test.go:313: 
            	Error Trace:	/home/runner/work/opentelemetry-collector-contrib/opentelemetry-collector-contrib/pkg/translator/jaeger/jaegerproto_to_traces_test.go:313
            	Error:      	Not equal: 
            	            	expected: ptrace.Traces{orig:(*v1.ExportTraceServiceRequest)(0xc000013470), state:(*internal.State)(0xc0001f2e2c)}
            	            	actual  : ptrace.Traces{orig:(*v1.ExportTraceServiceRequest)(0xc000013668), state:(*internal.State)(0xc0001f318c)}
            	            	
            	            	Diff:
            	            	--- Expected
            	            	+++ Actual
            	            	@@ -70,3 +70,3 @@
            	            	         ParentSpanId: (data.SpanID) (len=8) {
            	            	-         00000000  00 00 00 00 00 00 00 00                           |........|
            	            	+         00000000  af ae ad ac ab aa a9 a8                           |........|
            	            	         },
            	Test:       	TestProtoToTraces/two-spans-with-follower
    --- FAIL: TestProtoToTraces/a-spans-with-two-parent (0.00s)
        jaegerproto_to_traces_test.go:313: 
            	Error Trace:	/home/runner/work/opentelemetry-collector-contrib/opentelemetry-collector-contrib/pkg/translator/jaeger/jaegerproto_to_traces_test.go:313
            	Error:      	Not equal: 
            	            	expected: ptrace.Traces{orig:(*v1.ExportTraceServiceRequest)(0xc000013488), state:(*internal.State)(0xc0001f2e4c)}
            	            	actual  : ptrace.Traces{orig:(*v1.ExportTraceServiceRequest)(0xc0000138a8), state:(*internal.State)(0xc00022a84c)}
            	            	
            	            	Diff:
            	            	--- Expected
            	            	+++ Actual
            	            	@@ -70,3 +70,3 @@
            	            	         ParentSpanId: (data.SpanID) (len=8) {
            	            	-         00000000  00 00 00 00 00 00 00 00                           |........|
            	            	+         00000000  af ae ad ac ab aa a9 a8                           |........|
            	            	         },
            	Test:       	TestProtoToTraces/a-spans-with-two-parent
FAIL
	github.com/open-telemetry/opentelemetry-collector-contrib/pkg/translator/jaeger	coverage: 95.4% of statements
FAIL	github.com/open-telemetry/opentelemetry-collector-contrib/pkg/translator/jaeger	0.378s
FAIL

@frzifus @yurishkuro it looks like maybe this change: jaegertracing/jaeger@0ee275a, any thoughts on how this test should change to address the failure?

[yurishkuro]

@codeboten yes, that's the change. It allows a lone follows-from reference to be treated as a parent. However, the OTLP span that the Jaeger span is compared against after translation does not define parent ID. The tests can be fixed with this change:

diff --git a/pkg/translator/jaeger/jaegerproto_to_traces_test.go b/pkg/translator/jaeger/jaegerproto_to_traces_test.go
index 3ea43570d8..a75be85c87 100644
--- a/pkg/translator/jaeger/jaegerproto_to_traces_test.go
+++ b/pkg/translator/jaeger/jaegerproto_to_traces_test.go
@@ -919,6 +919,7 @@ func generateTracesTwoSpansWithFollower() ptrace.Traces {
        span.SetName("operationC")
        span.SetSpanID([8]byte{0x1F, 0x1E, 0x1D, 0x1C, 0x1B, 0x1A, 0x19, 0x18})
        span.SetTraceID(spans.At(0).TraceID())
+       span.SetParentSpanID(spans.At(0).SpanID())
        span.SetStartTimestamp(spans.At(0).EndTimestamp())
        span.SetEndTimestamp(spans.At(0).EndTimestamp() + 1000000)
        span.SetKind(ptrace.SpanKindConsumer)

However, it would cause breaks in some other conversion tests, for the same reason - one side does not include the reference and/or ParentID.

[codeboten]

@yurishkuro after making the suggested change, the following error occurs, will fixing this require a change in ThriftToTraces to support this as well?

                Error Trace:	opentelemetry-collector-contrib/pkg/translator/jaeger/jaegerthrift_to_traces_test.go:135
        	Error:      	Not equal: 
        	            	expected: ptrace.Traces{orig:(*v1.ExportTraceServiceRequest)(0x1400000f0c8), state:(*internal.State)(0x14000011d68)}
        	            	actual  : ptrace.Traces{orig:(*v1.ExportTraceServiceRequest)(0x1400000f290), state:(*internal.State)(0x14000011e98)}
        	            	
        	            	Diff:
        	            	--- Expected
        	            	+++ Actual
        	            	@@ -70,3 +70,3 @@
        	            	         ParentSpanId: (data.SpanID) (len=8) {
        	            	-         00000000  af ae ad ac ab aa a9 a8                           |........|
        	            	+         00000000  00 00 00 00 00 00 00 00                           |........|
        	            	         },
        	Test:       	TestThriftBatchToInternalTraces/a-spans-with-two-parent

[yurishkuro]

Yes, I believe translation code is incorrect. For instance, in makeJaegerProtoReferences, the parentID is unconditionally translated into CHILD_OF reference, but it should be conditional on not having a link with the same parentID, i.e.

1. process all links and convert them to references
2. if neither of them mentioned parentID, then prepend a CHILD_OF reference with that ID

Probably a similar issue in Thrift translation.