Add support for keepalive request time

[grcevski]

When a client is using keepalive for a request connection, the Go http server code will keep the goroutine live since the connection doesn't drop. This slightly defeats our request time vs. service time tracking, because there might be delay on processing the request after new data arrives on the old connection.

To mitigate this, we add a probe on startBackroundRead which is the code that will process the incoming data. This function is called very soon after the connection sees new data, so we'll be able to capture the full request time again, even in keepalive mode.

• Check what happens with Gorilla/mux
• Check what happens with Gin

[codecov-commenter]

Codecov Report

Merging #82 (2ceff1a) into main (dd00693) will decrease coverage by 0.16%.
The diff coverage is 50.00%.

@@            Coverage Diff             @@
##             main      #82      +/-   ##
==========================================
- Coverage   46.74%   46.58%   -0.16%     
==========================================
  Files          15       15              
  Lines        1166     1172       +6     
==========================================
+ Hits          545      546       +1     
- Misses        572      577       +5     
  Partials       49       49              

Flag	Coverage Δ
unittests	46.58% <50.00%> (-0.16%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
pkg/ebpf/nethttp/bpf_debug_bpfel_x86.go	0.00% <0.00%> (ø)
pkg/ebpf/nethttp/nethttp.go	0.00% <0.00%> (ø)
pkg/pipe/config.go	48.48% <ø> (ø)
pkg/pipe/instrumenter.go	80.00% <100.00%> (+0.37%)	⬆️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[grcevski]

Confirmed it works equally well with mux and gin. gRPC should be good, because as soon as the call ends it should terminate the connection. gRPC has it's own streaming long connection protocol.

[grcevski]

Thanks Mario!