CLI: get coredumps with request

[srgkvlv]

As an Operations Engineer, I want to be able to easily collect debug information to troubleshoot crashes and forward to support. I want to collect debug information from my work machine without having to shell the node running the Tarantool instance.

Acceptance Criteria:

1. The administrator can view the list of instances on a specific node, and for each instance, see information about the coredump on it:
   1. Number of coredump
   2. Serial number of the coredump
   3. The time of its appearance.
2. The administrator has the ability to download to his machine all coredump from all instances from a specific node with one command. If there is more than 1 coredump on a node (in total for all instances), then:
   1. Before loading the coredump, the administrator in the interactive mode is warned about the number of coredump and the total size of the files
   2. The administrator can agree to continue downloading or refuse to continue.
3. The administrator can download to his machine all the coredump of a particular instance with one command. If there is more than 1 coredump for this instance, then confirmation is requested in the same way as in p.
4. The administrator can download a specific coredump of a specific instance to his machine. To do this, he needs to specify the instance and the sequence number of the coredump.
5. To download to the administrator's machine, coredump and related artifacts are packaged as follows:
   1. Each coredump and its corresponding logs are packed into tar.gz archive
   2. If several coredumps are downloaded, then the archives of each coredump are packed into one common archive.
6. The administrator has the ability to delete all coredump from all host instances, all coredump from a specific instance, or a specific coredump of a specific instance (by analogy with unloading). Confirmation is requested from the administrator before deletion.
7. If the administrator requests to download or delete a non-existent coredump, he gets an error.