[Metricbeat] RabbitMQ module - not working with version 3.10

Metricbeat can not collect event.dataset "rabbitmq.node" for RabbitMQ v3.10.

Metricbeat returns error.message:
```
error in mapping: error applying schema: 3 errors: key `queue_index_journal_write_count` not found; key `io_file_handle_open_attempt_avg_time` not found; key `io_file_handle_open_attempt_count` not found
```

Using curl request (don't forget to change hostname to real hostname)

`curl -u guest -X GET http://localhost:15672/api/nodes/rabbit@hostname | jq '.|keys'`

we get all metric keys

```
[
"applications",
"auth_mechanisms",
"channel_closed",
"channel_closed_details",
"channel_created",
"channel_created_details",
"cluster_links",
"config_files",
"connection_closed",
"connection_closed_details",
"connection_created",
"connection_created_details",
"context_switches",
"context_switches_details",
"contexts",
"db_dir",
"disk_free",
"disk_free_alarm",
"disk_free_details",
"disk_free_limit",
"enabled_plugins",
"exchange_types",
"fd_total",
"fd_used",
"fd_used_details",
"gc_bytes_reclaimed",
"gc_bytes_reclaimed_details",
"gc_num",
"gc_num_details",
"io_read_avg_time",
"io_read_avg_time_details",
"io_read_bytes",
"io_read_bytes_details",
"io_read_count",
"io_read_count_details",
"io_reopen_count",
"io_reopen_count_details",
"io_seek_avg_time",
"io_seek_avg_time_details",
"io_seek_count",
"io_seek_count_details",
"io_sync_avg_time",
"io_sync_avg_time_details",
"io_sync_count",
"io_sync_count_details",
"io_write_avg_time",
"io_write_avg_time_details",
"io_write_bytes",
"io_write_bytes_details",
"io_write_count",
"io_write_count_details",
"log_files",
"mem_alarm",
"mem_calculation_strategy",
"mem_limit",
"mem_used",
"mem_used_details",
"metrics_gc_queue_length",
"mnesia_disk_tx_count",
"mnesia_disk_tx_count_details",
"mnesia_ram_tx_count",
"mnesia_ram_tx_count_details",
"msg_store_read_count",
"msg_store_read_count_details",
"msg_store_write_count",
"msg_store_write_count_details",
"name",
"net_ticktime",
"os_pid",
"partitions",
"proc_total",
"proc_used",
"proc_used_details",
"processors",
"queue_created",
"queue_created_details",
"queue_declared",
"queue_declared_details",
"queue_deleted",
"queue_deleted_details",
"queue_index_read_count",
"queue_index_read_count_details",
"queue_index_write_count",
"queue_index_write_count_details",
"ra_open_file_metrics",
"rates_mode",
"run_queue",
"running",
"sockets_total",
"sockets_used",
"sockets_used_details",
"type",
"uptime"
]
```

As you can see 3 metrics are missing:
- queue_index_journal_write_count
- io_file_handle_open_attempt_avg_time
- io_file_handle_open_attempt_count

Cause of this issue event.dataset "rabbitmq.node" doesn't return any metrics at all.

According to https://github.com/rabbitmq/rabbitmq-server/pull/4431 metrics were deleted cause been proven irrelevant.

There is a PR for RabbitMQ where some disk metrics were made optional https://github.com/elastic/beats/pull/6887 may be this ones should be too.