Changes for multiple OSDs and aws ec2

[ZhaoNeil]

Call cmake twice due to the crash the first time.

[ZhaoNeil]

Main contribution of this pull request: changes for multiple OSDs deployments in a single NVMe and AWS EC2 configuration.

In particular:

1. Change the script for multiple OSDs deployments in a single NVMe using the command ceph-volume lvm batch --osds-per-device.
2. Local area network configuration for AWS cluster. Previously script is for cloudlab, where the nodes can be configured as small-LAN when reserved. But for AWS, it needs the local area network configuration to make sure CEPH cluster work well.

Additionally:

1. Add simplified Jayjeet's script to reserve AWS nodes automatically.
2. Fix the bug where we need call cmake twice.