大数据集群（Hadoop，Spark等）中的副本是否由同一系统管理员管理？

Question

I am a security researcher and would like to know the standard practice about managing replicas in a big data cluster. Specifically, I would like to know if platforms like Hadoop, Spark etc [or] providers like Google, Amazon etc usually employ the same sysadmin to manage replica nodes? If so, will it affect the performance so bad if the replicas are managed by different sysadmins?

I know that the scope of this question is too broad, so any insight into big data security standards about replica management is appreciated. Thanks.

Answer 1

The sysadmin of a slave node does not matters as most of the communication within the cluster happens over the network. And in my view sysadmin cannot impact the performance of a Hadoop Cluster.