Unable To Restore From Backup Due To Ruby Error

[rkotecki]

Describe the problem

I am getting a generic ruby syntax error when attempting a restore from a backup. Below is a snippet of the restore log with the failure message. I have also attached more of the log further down
``
Copying /cookbook_artifacts
FATAL: Cookbook file recipes/filebeat.rb has a ruby syntax error:

I checked the backup and there are 530 instances of recipes/filebeat.rb.

[root@chef-server backup]# find -name filebeat.rb | wc -l
530

I would think that it wouldn't let me do a `knife ec backup` if there is a ruby error that won't let me import via `knife ec restore`

## Software Version
[root@chef-server backup]# chef -v
ChefDK version: 4.10.0
Chef Infra Client version: 15.13.8
Chef InSpec version: 4.22.1
Test Kitchen version: 2.5.4
Foodcritic version: 16.3.0
Cookstyle version: 5.23.0
[root@chef-server backup]# chef-server-ctl version
13.2.0

## Stacktrace

Copying /acls/containers
Copying /acls/environments
Copying /acls/organization.json
Copying /groups/admins.json
Copying /groups/billing-admins.json
Copying /groups/public_key_read_access.json
Restoring organization[dfs-qa]
/root/.chefdk/gem/ruby/2.6.0/gems/knife-ec-backup-2.4.7/lib/chef/knife/ec_restore.rb:75: warning: constant Net::HTTPServerException is deprecated
/root/.chefdk/gem/ruby/2.6.0/gems/knife-ec-backup-2.4.7/lib/chef/knife/ec_restore.rb:105: warning: constant Net::HTTPServerException is deprecated
Restoring org admin data
Copying /groups/admins.json
Copying /groups/billing-admins.json
Copying /groups/public_key_read_access.json
Copying /acls/groups/billing-admins.json
Copying /acls/groups/public_key_read_access.json
Restoring the rest of the org
Copying /clients
Copying /containers
Copying /cookbook_artifacts
FATAL: Cookbook file recipes/filebeat.rb has a ruby syntax error:
/root/.chefdk/gem/ruby/2.6.0/gems/knife-ec-backup-2.4.7/lib/chef/knife/ec_restore.rb:320: warning: constant Net::HTTPServerException is deprecated

[devopsdina]

@rkotecki Can you try the following:

Verify the chef server is in a good working state prior to taking the backup:

• chef-server-ctl test

Can you confirm the backup is being taken from an isolated ruby environment (i.e. chefdk or chef-workstation) with a concurrency of 1:

• /opt/chefdk/embedded/bin/knife ec backup my_backup_destination --with-user-sql --with-key-sql --concurrency 1 -c /etc/opscode/pivotal.rb

Can you try running a knife-tidy against the backup:

• /opt/chefdk/embedded/bin/knife tidy backup clean --backup-path my_backup_destination

If the restore is happening on a new OS (migration) vs. an in-place upgrade, ensure the newly provisioned OS has the proper libraries installed:
For Ubuntu:

• sudo apt-get -y install gcc postgresql libpq-dev

For RHEL/Centos:

• sudo yum -y install gcc postgresql-devel