Error rate / Errors in trace

[danibachar]

Hello there, First, thank you very much for releasing such a great data set!

I was wondering if you have recorded the error rates (i.e requests the pod could not answer due to overload or internal error) from each pod/sidecar? If so is there an option to get this data as well?

[qzweng]

Hi @danibachar,

Could you kindly point out which trace/data set you are referring? So that the maintainer of that dataset may respond when available.

[danibachar]

Thanks @qzweng, I appreciate your response and apologies for the delay in mine.

I referred to the traces in cluster-trace-microservices-v2021, and specifically to the following tables

• MS_MCR_RT_Table
• MS_CallGraph_Table

Please note that I'm well aware that there are no error rates/information in those traces. I was just wondering if those are statistics you collect from the sidecar proxy at each container(for example 500/400 HTTP errors) and can share them in the same traces or others. Those would be very helpful in identifying patterns.

Or maybe the error rate is so insignificant that is not worth mentioning?

[niewuya]

The error rate can reflect the quality of services somehow. It will be useful for research especially in computer networks. We prepare to release the new microservices-trace-v2022 in the coming days. We consider releasing the error rate as a new metric. Please keep looking at the new dataset.

[danibachar]

Thank you very much @niewuya . We would very much appreciate the release with the error rate!