Add ability to enforce SSL in Databricks Destination connector

[ryankfu]

Tell us about the problem you're trying to solve

Databricks does not always enforce SSL

Describe the solution you’d like

Implement configuration for adding ssl parameter for Databricks destination

Describe the alternative you’ve considered or used

A clear and concise description of any alternative solutions or features you've considered or are using today.

Additional context

Building the connection URL

Are you willing to submit a PR?

Remove this with your answer :-)

[grishick]

cc @tuliren

[grishick]

Notes on scope:

• figure out if DataBricks allows unencrypted connections (if it doesn't, this is a no-op 🎉 ). Otherwise:
  • add SSL on/off switch to spec
  • enforce SSL in Airbyte cloud unless SSH tunnel is on (this includes normalization)
  • understand and expose (if available) SSL modes for DataBricks JDBC driver. If various SSL modes are available, allow only secure modes in Airbyte Cloud
  • add tests that verify the above behaviors

[grishick]

Hey team! Please add your planning poker estimate with Zenhub @edgao @rodireich @ryankfu @subodh1810 @tuliren

[grishick]

Investigation

I plaid around with our databricks sandbox account and found the following:

• Databricks clusters always require SSL to be enabled in JDBC connections. I tried disabling it and was not able to connect
• If JDBC connection URL does not explicitly enable SSL, SSL is still enabled by default
• If JDBC connection URL explicitly disables SSL, Databricks cluster refuses the connection

Here are some references:

• DBeaver (or any other DB browser) integration
• JDBC configuration guide

Remaining scope for this issue

I also found that our JDBC connection string for Databricks currently does not conform to the JDBC connection guide and is missing ssl=1 parameter. Although this parameter is implied, we should still add it as recommended in Databricks documentation. My understanding is the only thing we need to do is to add ssl=1 JDBC URL parameter hardcoded in Databricks connector per Databricks documentation. The reason it has to be hardcoded is because, Databricks does not allow any value for this parameter.

Changing the estimate to 2 points based on these findings.

[grishick]

Moving to Blocked until @tuliren finishes current Databricks changes