Enhance CI benchmarks with regression detection

[kamilkisiela]

This commit introduces a new script to detect performance regressions. It compares the 'http_reqs' rate metric from the PR summary against the main branch summary and determines if there is a regression of more than 2%.

Also, the commit adds .gitignore entries, a script to detect regressions, and changes to the CI workflow to execute the benchmarks and compare the results.

[github-actions]

🐋 This PR was built and pushed to the following Docker images:

Image Names: ghcr.io/graphql-hive/router

Platforms: linux/amd64,linux/arm64

Image Tags: ghcr.io/graphql-hive/router:pr-531 ghcr.io/graphql-hive/router:sha-52abb5a

Docker metadata

{
"buildx.build.ref": "builder-a45b52bd-33cb-4549-99e9-7def76a6d499/builder-a45b52bd-33cb-4549-99e9-7def76a6d4990/97w5vilsbswt2552d4mps7185",
"containerimage.descriptor": {
  "mediaType": "application/vnd.oci.image.index.v1+json",
  "digest": "sha256:a7b181d78ec616d8c5279c1916fc135fdf3f6394b1b1b5a8d4d0b2fa5adf7f60",
  "size": 1609
},
"containerimage.digest": "sha256:a7b181d78ec616d8c5279c1916fc135fdf3f6394b1b1b5a8d4d0b2fa5adf7f60",
"image.name": "ghcr.io/graphql-hive/router:pr-531,ghcr.io/graphql-hive/router:sha-52abb5a"
}

[github-actions]

✅ k6-benchmark results

     ✓ response code was 200
     ✓ no graphql errors
     ✓ valid response structure

     █ setup

     checks.........................: 100.00% ✓ 226365      ✗ 0    
     data_received..................: 6.6 GB  220 MB/s
     data_sent......................: 89 MB   2.9 MB/s
     http_req_blocked...............: avg=3.11µs   min=639ns   med=1.65µs  max=11.54ms  p(90)=2.39µs  p(95)=2.77µs  
     http_req_connecting............: avg=647ns    min=0s      med=0s      max=2.07ms   p(90)=0s      p(95)=0s      
     http_req_duration..............: avg=19.45ms  min=2.06ms  med=18.53ms max=95.78ms  p(90)=26.67ms p(95)=29.73ms 
       { expected_response:true }...: avg=19.45ms  min=2.06ms  med=18.53ms max=95.78ms  p(90)=26.67ms p(95)=29.73ms 
     http_req_failed................: 0.00%   ✓ 0           ✗ 75475
     http_req_receiving.............: avg=131.82µs min=27.74µs med=44.61µs max=62ms     p(90)=79.96µs p(95)=368.81µs
     http_req_sending...............: avg=20.64µs  min=4.34µs  med=10.15µs max=18.68ms  p(90)=14.75µs p(95)=20.77µs 
     http_req_tls_handshaking.......: avg=0s       min=0s      med=0s      max=0s       p(90)=0s      p(95)=0s      
     http_req_waiting...............: avg=19.29ms  min=2ms     med=18.41ms max=73.42ms  p(90)=26.43ms p(95)=29.46ms 
     http_reqs......................: 75475   2509.870296/s
     iteration_duration.............: avg=19.87ms  min=5.63ms  med=18.85ms max=228.91ms p(90)=27.06ms p(95)=30.17ms 
     iterations.....................: 75455   2509.205209/s
     vus............................: 50      min=50        max=50 
     vus_max........................: 50      min=50        max=50 

[dotansimha]

we do not chain PR a lot, but this should really be ${{ github.base_ref }}