Some Questions about AMTrace

1. I noticed that nearly half of the containers only have 'node_id' and 'container_id' field infomation. I wonder if it is convenient for you to tell us why other fileds are missing.
2. In container_meta.csv, some pod request <= container request. As far as I know, it is reasonable to see pod request >= container request but it is wierd to see pod request <= container request. I wonder in what situation could happen pod request <= container request.
3. The files in the 'core_pmu_1k/' directory only have 10 fileds but README said that there should be 11 fields. I looked at the files and concluded that 'pod_id' is not in those files. I think maybe you could rectify the README.
4. There exists some container_id appearing multiple times in the 'container_meta.csv'. They have the same container_id but their node_ids are different. I wonder if it is the situation that some tasks are running on multiples nodes and the same container id on different nodes means that these nodes are running the same task.
I would be grateful if anyone can response to the above questions.

Start with the README, container_meta.csv, and the files under core_pmu_1k/. Compare the documented schemas and field values, then trace the dataset definitions or collection notes needed to explain missing fields, request relationships, and repeated container IDs. Done means the questions have documented answers and the README accurately lists the available fields.

Written by the indexing model from the issue text.