benchmark(bigtable): update sync and async scan benchmarks #15557
Conversation
product-auto-label
bot
added
the
api: bigtable
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #15557 +/- ##
=======================================
Coverage 93.03% 93.04%
=======================================
Files 2408 2408
Lines 220145 220149 +4
=======================================
+ Hits 204810 204833 +23
+ Misses 15335 15316 -19 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
| #ifdef PROFILE | ||
| auto profile_data_path = google::cloud::internal::GetEnv("PROFILER_PATH"); | ||
| if (profile_data_path) ProfilerStart(profile_data_path->c_str()); | ||
| auto profiler_start = std::chrono::steady_clock::now(); | ||
| #endif // PROFILE |
There was a problem hiding this comment.
not a blocker for the PR but maybe as a follow-up: it might be worth adding documentation to the README or onboarding docs on how to perform throughput benchmarking with this newly added logic?
There was a problem hiding this comment.
Added doc link and execution instructions.
| {"--enable-metrics", | ||
| "whether to enable Client Side Metrics for benchmarking", | ||
| [&options](std::string const& val) { | ||
| options.enable_metrics = ParseBoolean(val).value_or(true); | ||
| }}, |
There was a problem hiding this comment.
would you provide some context into why we are choosing to have client-side metrics for benchmarking behind a flag?
There was a problem hiding this comment.
Because some of the latency metrics are part of the inner loop of delivering assembled rows to the application, I want to be able to easily compare the performance impact of metrics on versus off.
There was a problem hiding this comment.
Thanks for the context!
scotthart
left a comment
scotthart
left a comment
There was a problem hiding this comment.
Reviewable status: 0 of 8 files reviewed, 2 unresolved discussions (waiting on @mpeddada1)
| {"--enable-metrics", | ||
| "whether to enable Client Side Metrics for benchmarking", | ||
| [&options](std::string const& val) { | ||
| options.enable_metrics = ParseBoolean(val).value_or(true); | ||
| }}, |
There was a problem hiding this comment.
Because some of the latency metrics are part of the inner loop of delivering assembled rows to the application, I want to be able to easily compare the performance impact of metrics on versus off.
| #ifdef PROFILE | ||
| auto profile_data_path = google::cloud::internal::GetEnv("PROFILER_PATH"); | ||
| if (profile_data_path) ProfilerStart(profile_data_path->c_str()); | ||
| auto profiler_start = std::chrono::steady_clock::now(); | ||
| #endif // PROFILE |
There was a problem hiding this comment.
Added doc link and execution instructions.
2 participants
Adds scan_async_throughput_benchmark as an async version of scan_throughput_benchmark. Both now support profiling around ReadRows sections which require gperftools to be installed and can be enabled via:
This change is