Give all users read/write permissions for snapshots directory

[caffeinatedpixel]

Closes #80

In order to resolve the missing permissions on the snapshots directory for some BeaKer installs, the directory is given read/write permissions for all users. This shouldn't be an issue since the /opt/BeaKer directory requires admin/sudo access.

Testing:

I tried to replicate the issue on Ubuntu 20.04 with a fresh v0.0.13 BeaKer install and applying the current installer over it for ELK v7.17.9 and v8.7.0. I did not run into the failing snapshot creation during the upgrade. After the upgrade completed, manually created a snapshot via the snapshot API and was able to do so. I exec'd into the elasticsearch container and noted that the owner of the snapshots directory was elasticsearch:root, and was able to write to that directory via echo "test" > test. The /opt/BeaKer/snapshots folder was owned by naomi:root. Creating a file in /opt/BeaKer/snapshots created a file in /usr/share/elasticsearch/snapshots within the container. After running beaker down and beaker up -d, the permissions stayed the same.
I repeated the steps on CentOS 7 ( selinux mode permissive) with the same results.

I tested this PR using the same steps listed above on Ubuntu 20.04. Since the folder ownership change was removed, the owner is now root:root for both /opt/BeaKer/snapshots and /usr/share/elasticsearch/snapshots on the container. In my case, snapshot creation continues to succeed.

[Zalgo2462]

I built this installer and tried to install it on my Ubuntu 20 machine which has the issue.

I'm not able to make a snapshot but it appears the error may be different at first glance:

Running the snapshot repo creation command again yields the same 500 error message:

The elasticsearch logs have the same errors:
org.elasticsearch.repositories.RepositoryVerificationException: [beaker] path is not accessible on master node
Caused by: java.nio.file.NoSuchFileException: /usr/share/elasticsearch/snapshots/tests-7TLlg7t1QP6UC6KVtjd_Ww

A few things I noticed:

The script attempts to restart beaker even though it is already running.
The following line should use a dash or wildcard between beaker and elasticsearch:

BeaKer/installer/stage/BeaKer/install_beaker.sh

Line 294 in c2efade

if [ `$SUDO docker ps | grep 'beaker_elasticsearch' | wc -l` -eq 0 ]; then

My docker ps output:

We don't mark the host directory 777 at https://github.com/activecm/BeaKer/blob/master/installer/stage/BeaKer/install_beaker.sh#L181 . I'll try adding a chmod to this line here and giving it another run.

[Zalgo2462]

I'm going to run a test with the following changes:

Let me know if you have any other ideas.

[Zalgo2462]

The changes in the diff above added to this PR worked to get this machine to take a snapshot.
@caffeinatedpixel do those changes look good to you?

[caffeinatedpixel]

@Zalgo2462 Looks good, pushed

[Zalgo2462]

@caffeinatedpixel The chmod line in ensure_snapshot_repo_exists needs to be added as well.

[caffeinatedpixel]

@Zalgo2462 My bad, added as well

[Zalgo2462]

Tested well