Move snapshot directory to /var/BeaKer/snapshots

[caffeinatedpixel]

Closes #82

The /opt/BeaKer directory is deleted upon every run of the installer. This means that the snapshots subdirectory and its files are also deleted. This PR moves the snapshots directory to /var/BeaKer/snapshots, which would not get deleted on every install.

Tested on Ubuntu 20.04 & CentOS 7

• Generated new BeaKer installer
• Installed new installer over a BeaKer v0.0.13 install
• Ran ./beaker down && beaker up -d for sanity
• Ran the installer again to upgrade to Elastic v8.7.0 and created a snapshot
• Verified that snapshot was created and that the snapshot files were visible in /var/BeaKer/snapshots & /usr/share/elasticsearch/snapshots
• Ran the installer again to overwrite the v8.7.0 install, verified that the snapshot files were still in the host & container directories

[Zalgo2462]

I tested this PR using the procedure above. During the reinstall of 8.x over itself, I got passed the old error, but I ran into a new error

Looking at my env file, the token is present, so I'm not sure why the script did this.

Looking at the script, it appears token_already_valid was false for some reason.

The script attempted to hit the Elastic API with the service token. For some reason, this failed.
Then, the script tried to check if a token already existed using the elastic password. For some reason, this failed.
Then, the script tried to create a new service token, but Elastic returns an error if you try to create a new a token without deleting the old one. As a result, the script crashed.

Perhaps Elastic wasn't really ready by the time the API was called? Running these commands on the cli afterwards seems to work.

[Zalgo2462]

Running the script again, I did not run into this issue. The problem seems to be solved by re-running the installer. So, I'm going to pull this in.