How to code instance name before dataset of cluster-trace-v2018 published.

[marsjiang216]

Dear Mr. Ding,

Hello! I am a graduate student. We are very interested in the "Alibaba Cluster-trace-V2018" dataset and have downloaded the dataset.

We found that the instance name was formatted as "<ins><_><digital>, for example,ins_26041377" where the last four digits of the string are different and the first part of the string has several categories when we interpreted the dataset.

At the same time, we looked at some help documents on aliyun, where instance names are encoded in the format "<task_name><#><_>, for example, R2_4#281_0".

We want to know how you handle it encoding from instance A("R4_2#281_0") to instance A'("ins_26041377") before you publish this dataset. What are the mapping rules?

We are looking forward to your reply!

Sincerely Yours
Marsjiang, 20201012