feat(duplication): add options to support cluster name only used for duplication and allow any other cluster id except myself to be ignored

[empiredan]

The purpose of this PR is to optimize configurations for duplications.

Firstly, many Pegasus clusters are configured with the same cluster_name
(namely [replication]cluster_name). However, once we decide to duplicate
tables between them, their cluster_name have to be changed to be distinguished
from each other -- this might lead to side effects.

Secondly, consider a scenario where many clusters are duplicated to a target
cluster. This means we have to add many cluster ids to the *.ini file of the target
cluster, and the target cluster might be restarted very frequently.

Thus following options are added to solve both problems:

[replication]
+ dup_cluster_name =
+ dup_ignore_other_cluster_ids = false

[replication]dup_cluster_name is added only for duplication in case cluster_name
has to be changed, while [replication]dup_ignore_other_cluster_ids is added so that
only the target cluster id should be configured and there is no need to add any other
cluster id.

[acelyc111]

Thumbs for the 2000th PR/issue!