Orderer addition job issues

[jithindevasia]

• The job should get exited immediately if any step fails.
• A new check should be added in the job beginning itself to fetch config blocks for all listed channels and check if there is an entry for this new orderer exists already. If an entry entry exists, then it should not even try anything and rather print a message to user for a manual removal of this orderer entry from the channel and re-run.
• Add strict validation and error handling for tls cert tar generation and genesis block upload to filestore. Because if TLS certs are gone, then there is no way to recover it again.
• There are chances that, while adding an orderer to multiple channels one may success and other one may fail. In this case, re-running the same job will create another set of key pairs and update second channel. So the same orderer will be having two sets of TLS cert entries in two channels which SHOULD NOT happen. Job should handle this as well.

Affected version Fabric-ops 1.1.0