Inconsistency in Microservice Dataset Across Different Tables (cluster-trace-microservices-v2022)

[Alireza-Ezaz]

Hello @niewuya ,

I am currently working with the 1-hour microservices dataset released by your team, which includes the following tables: CallGraph_0, MCRRTUpdate_0, MSMetricsUpdate_0, and NodeMetricsUpdate_0.

I have encountered a challenge when attempting to connect these tables in a meaningful way. Specifically, there seems to be an inconsistency with the msinstanceid values between the CallGraph_0 and MSMetricsUpdate_0 tables.

For example, in CallGraph_0, there is a row with traceid T_13736926193 where um is MS_40586 and uminstanceid is MS_40586_POD_1010. When I attempt to match this to the MSMetricsUpdate_0 table to find the corresponding metrics for um MS_40586, I find records where the msname matches um, but the msinstanceid does not match the uminstanceid from CallGraph_0.

I am looking for guidance on how to interpret these relations. Is there a linking key that I am missing, or is this an expected discrepancy due to the nature of the data?