neo4j-admin 3.5 backup does not explain it could not connect to server

Guidelines

To help us understand your issue, please specify important details, primarily:

ONgDB version: 3.5.17
Operating system: Linux (Ubuntu 18.04)
API/Driver: (for example Cypher/Java API/Python driver vX.Y.Z)

Steps to reproduce

Do not run a ongdb server (or any other server) on 127.0.0.1:6462

Ask neo4j-admin to perform a backup from that address:

ongdb_backup=$(mktemp -d)
report_dir=$(mktemp -d)
JAVA_HOME=$(oracle_jdk8) \
neo4j-admin backup \
    --protocol=catchup \
    --from=127.0.0.1:6462 \
    --backup-dir=$ongdb_backup \
    --cc-report-dir=$report_dir \
    --name=backup; echo $?

Expected behavior

An error message indicating that no server responded from `127.0.0.1:6462`.
Bonus points for an exit code that distinguishes between "spoke to a server" and "couldn't speak to a server".

Actual behavior

```
The selected protocol `catchup` means that it is only compatible with causal clustering instances
command failed: Failed to run a backup using the available strategies.
1
```

Additionally, include (as appropriate) log-files, stacktraces, and other debug output.

There are no log files generated -- this is part of the problem
There are no stack traces generated -- this is part of the problem
There is no debug output, and there isn't any obvious way to enable debug output -- this is part of the problem (version 4.0 supports a --verbose flag -- I'm not sure if it would help)

https://neo4j.com/docs/operations-manual/3.5/backup/performing/#backup-performing-command says that 1 means:

Backup failed.

This is correct, the backup failed.

It also isn't helpful. It'd be much more helpful if I could distinguish between "could not connect to server" and "I spoke to the server and things went bad". I'm fairly certain that the second case can happen. -- If not, the documentation should just say that 1 means "could not connect to server" -- and again, I highly doubt that's what it means.

graphfoundation / ongdb