Note that this feature will be enabled by default on all Spotlight Cloud accounts starting from December of 2019. Users can opt to disable this feature in the on-premise configuration tool after deployment if required.
Spotlight Cloud is responsible for performance monitoring. As a result, it collects tons of data about the health and performance of a monitored database. It’s important to note that it does not actually access the data in the database but instead queries sources of information such as DMVs to establish what was happening, how it happened, and when it happened.
One of the most popular features is to show Top SQL. Spotlight Cloud is constantly monitoring the SQL statements that consume most of your workload. By examining the SQL code, the associated wait statistics, and the plan that was used to execute the statement, the user can gain a detailed view of how they might improve the performance of the code.
Well-written code in a production environment with sensitive data will use bind variables to improve performance and security. This prevents any “data fragments” being embedded in SQL by way of variables or literals. If coding standards are poorly enforced or someone executes some ad-hoc query that uses unmasked literals or variables, these would get uploaded into Spotlight Cloud (or any similar monitoring tool) as it monitors Top SQL.
Without a doubt, Spotlight Cloud encrypts all data in transit and in storage. Your data is safe. However, like all things, problems can still arise even if proper measures are in place. For example, in a GDPR “Right to be forgotten” request from a customer, you may wish to scrub any trace of the user's data from any of your systems. It is possible that some trace may persist in logging or monitoring tools as a result of careless coding or ad-hoc queries.
In Spotlight Cloud, we give you the option to eliminate this risk. “Remove literals” is a feature that scans any SQL at the point of collection. The on-premise diagnostic server scans the code and then removes the literals before they are uploaded. This ensures that sensitive data fragments do not leave your organization.
In the case of string literal, the literal will be replaced with ‘**Removed by Spotlight **’. In the case of a numeric literal, the numeric is removed and replaced with placeholder numeric data. The reason for treating them differently is to keep the query in a state that is executable to aid performance tuning and diagnostics.
If you would rather keep your SQL completely intact, you can switch off literal screening in the on-premise configuration app.
Spotlight Cloud is committed to the highest standards of security and data governance. We are delighted to bring you this feature to help keep your data safe and remove the risk of any unintended non-conformities.