All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=754693

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=157cbd2a225c9b6877adf87c92ae805e271f31b35447372999f80420a740b178


Bot(s) for this bug's original alert(s):

chromium-rel-win8-dual

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8971552791639028864

=== BISECT JOB RESULTS ===
NO Perf regression found

Bisect Details
  Configuration: win_8_perf_bisect
  Benchmark    : tracing.tracing_with_debug_overhead
  Metric       : trace_size_avg/http___www.amazon.com

Revision             Result                    N
chromium@490927      10787755 +- 25301680      21      good
chromium@491274      14403352 +- 41392359      21      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=http...www.amazon.com tracing.tracing_with_debug_overhead

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8971552791639028864


For feedback, file a bug with component Speed>Bisection

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8968081830539616480

This benchmark has been turned down.

=== Auto-CCing suspected CL author xunjieli@chromium.org ===

Hi xunjieli@chromium.org, the bisect results pointed to your CL, please take a look at the
results.


=== BISECT JOB RESULTS ===
Perf regression found with culprit

Suspected Commit
  Author : xunjieli
  Commit : 357e1deefca78c61e457f340270c9cd6f38321f5
  Date   : Thu Aug 03 22:16:30 2017
  Subject: [wpr-go] Use a dummy cert if no cert is recorded

Bisect Details
  Configuration: win_8_perf_bisect
  Benchmark    : tracing.tracing_with_debug_overhead
  Metric       : trace_size_avg/http___www.amazon.com
  Change       : 60.19% | 10224601.4762 -> 16378818.8095

Revision                                 Result                    N
chromium@490844                          10224601 +- 11610601      21      good
chromium@491399                          12852643 +- 18077268      6       good
chromium@491676                          14317237 +- 25245017      6       good
chromium@491815                          13523066 +- 22956329      6       good
chromium@491884                          13319445 +- 19952778      6       good
chromium@491919                          12858539 +- 24132060      14      good
chromium@491924                          10050616 +- 5233290       9       good
chromium@491924,catapult@b0f1e24d46      11004978 +- 15416887      9       good
chromium@491924,catapult@357e1deefc      12601905 +- 15170398      9       bad       <--
chromium@491925                          14752495 +- 32513438      9       bad
chromium@491926                          13907311 +- 24133007      9       bad
chromium@491928                          12057575 +- 14060395      14      bad
chromium@491936                          12893210 +- 19860183      14      bad
chromium@491953                          16378819 +- 55570720      21      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=http...www.amazon.com tracing.tracing_with_debug_overhead

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8968081830539616480


For feedback, file a bug with component Speed>Bisection

Does "tracing.tracing_with_debug_overhead" measure the memory usage?
We are using more memory with WPRGo.

We may get 30% more trace events because they are emitted by network code that only exercised with WPRGO?

Re#8: That's very possible. Let me archive this. I don't think it's worth digging deeper given that this is an infra change and not a real regression.