Run FIO in multiple instances

I would like the ability to run FIO on multiple instances all starting at the same time as a stress test to a cloud.

Currently I can run:

```
$ ./pkb.py --cloud=OpenStack --openstack_private_network=osp_private --config_override=fio.vm_groups.default.vm_count=2 --image=centos7 --openstack_image_username=centos --data_disk_size=10 --fio_run_for_minutes=10 --os_type=rhel --benchmarks=fio --fio_generate_scenarios=all --machine_type=m1.small --fio_working_set_size=4 --openstack_public_network=osp_public
```

Although two instances are started, fio only runs on one.

The help text is a bit misleading as well:

```
$ ./pkb.py --help
PerfKitBenchmarker version: v1.2.0-48-g81c28f0
...
(variable VMs)
copy_throughput: Get cp and scp performance between vms. (1 VMs with scratch volume(s))
coremark: Run Coremark a simple processor benchmark (1 VMs)
fio: Runs fio in sequential, random, read and write modes. (1 VMs with scratch volume(s))
hadoop_terasort: Runs Terasort. Control the number of worker VMs with --num_vms. (9 VMs with scratch volume(s))
hbase_ycsb: Run YCSB against HBase. Specify the HBase cluster size with --num_vms. Specify the number of YCSB VMs with --ycsb_client_vms.
....
```

Why are the VMs variable above those benchmarks in the help output, yet only one vm will run fio. I am guessing the help output is not up to date?

Is multiple concurrent instances of FIO on the roadmap to be implemented?