alibaba / clusterdata

cluster data collected from production clusters in Alibaba for cluster management research
1.57k stars 405 forks source link

Microservices Paper (SoCC '21) Clarification #148

Open MaxLanLiu opened 2 years ago

MaxLanLiu commented 2 years ago

Hi I have a clarification question regarding a figure and its textual explanation in your paper, Characterizing Microservice Dependency and Performance: Alibaba Trace Analysis.

In Figure 3(a) (see the screenshot below), did you mean the number of unique microservices in a call graph? I guess you are not really counting "unique" ones.

Figure 3a

But in the corresponding textual explanation of Figure 3(a), "unique microservices" is used once.

text

To get Figure 3(a), if one of the call graphs is the following example in the picture, will the analysis count this call graph contain 2 unique microservices or 3 microservices?

example

Thanks again!

niewuya commented 2 years ago

Thanks for your interest! We count the unique microservices in Fig. 3(a). In the given example, the call graph contains two unique microservices.

andreadetti commented 2 years ago

Hi, I have another question about the paper. The Stocastic graph modeling makes use of the F_h(j) distribution, which is the distribution of the number of microservices in a two-level invocation starting from level h. Have you released this distribution somewhere? Thank you very much.

niewuya commented 2 years ago

Thanks for your interest! Please refer to Figure 8 in the paper for more information about tier-level microservice statistics.