Why does the caller of TaskExecutorGateway.requestThreadInfoSamples need to specify the request id? Is it because the caller can send a second request with the same id? Or can the caller query the result of a previous request by specifying the requestId?
If the TaskExecutor does not need to know about the id, then we might be able to drop it. Cheers Till On Tue, Mar 2, 2021 at 9:42 AM Alexander Fedulov <alexan...@ververica.com> wrote: > Hi Till, > > Thanks for your comments. > > * What is the requestId used for in the RPC call? > > It is the handle that is used as the key in the > ThreadInfoRequestCoordinator's pending responses Map. I believe it was > called sampleId in the StackTraceSampleCoordinator, but I decided to rename > it because there is also a ThreadInfoSampleService which is actually > responsible for sampling the JVM numSamples number of times. I found that > the notion of what a sample is was a bit confusing. Now one thread info > request corresponds to gathering numSamples from a corresponding Task. Hope > that makes sense. > > * Would it make sense to group numSubSamples, delayBetweenSamples and > maxStackTraceDepth into a ThreadSamplesRequest class? This would decrease > the number of parameters and group those which are closely related > together. > > Good point. I will rework it accordingly. > > Best, > -- > > Alexander Fedulov | Solutions Architect > Follow us @VervericaData > > > > -- > Sent from: http://apache-flink-mailing-list-archive.1008284.n3.nabble.com/ >
