Initial commit for distributed tests using crank #7323
Conversation
| var counterGrain = client.GetGrain<ICounterGrain>(parameters.CounterKey); | ||
|
|
||
| var duration = await counterGrain.GetRunDuration(); | ||
| BenchmarksEventSource.Register("duration", Operations.First, Operations.Last, "duration", "duration", "n0"); |
There was a problem hiding this comment.
A comment or named parameter might help to make it clear what "n0" means
| _logger.LogInformation($"{counter}: {value}"); | ||
| BenchmarksEventSource.Register(counter, Operations.First, Operations.Sum, counter, counter, "n0"); | ||
| BenchmarksEventSource.Measure(counter, value); | ||
| if (string.Compare(counter, "requests", StringComparison.InvariantCultureIgnoreCase) == 0) |
There was a problem hiding this comment.
string.Equals seems to be the right choice instead, but it doesn't matter
| <OutputType>Exe</OutputType> | ||
| <TargetFrameworks>$(TestTargetFrameworks)</TargetFrameworks> | ||
| <OrleansBuildTimeCodeGen>true</OrleansBuildTimeCodeGen> | ||
| <OutputPath Condition=" '$(DistributedOutputPath)'!='' ">$(DistributedOutputPath)/Distributed.Client</OutputPath> |
There was a problem hiding this comment.
"DistributedTestsOutputPath" would be clearer
| public async Task IssueRequest(int request) | ||
| { | ||
| var (router, targets) = _requestList[request % _requestList.Count]; | ||
| await router.Ping(targets); |
There was a problem hiding this comment.
For benchmarking purposes, it may make sense to not use 'await' and to elide these instead - but we can do that test optimization later
|
|
||
| public CounterGrain(IOptions<ReportingOptions> options) | ||
| { | ||
| Debugger.Break(); |
There was a problem hiding this comment.
I think we should remove this Debugger.Break() - I recall that it can cause issues in some environments, but maybe I'm misremembering.
| break; | ||
| } | ||
| } | ||
| while (_concurrencyGuard.CurrentCount != commonParameters.PipelineSize) |
There was a problem hiding this comment.
I think this technique will perform relatively poorly compared to what we have with ConcurrentLoadGenerator https://github.com/dotnet/orleans/blob/main/test/Benchmarks/Ping/ConcurrentLoadGenerator.cs
benjaminpetit
force-pushed
the
distributed-tests
branch
from
f7969ad to
f150852
Compare
| @@ -148,7 +148,7 @@ public void RecordDeliveryFailure() | |||
| /// <summary> | |||
| /// The limit of the maximum number of items that can be added | |||
| /// </summary> | |||
| public int GetMaxAddCount() { return 100; } | |||
| public int GetMaxAddCount() { return 200; } | |||
There was a problem hiding this comment.
This seems unrelated to the rest
null
Microsoft Reviewers: Open in CodeFlow