2017-2018 exchange Nov Dec 2017 Newsletter FINAL | Page 13

ees are not conducting appropriate disposition if they are not deleting structured data ?
BS : Its hard to say . Lets take this specific example : there are classes of records in the past that were maintained in a paper format , like accounting records , that now no longer exist in that format . So whatever retention period you had for those records , they would apply to the databases that have replaced them . It ’ s not always conveniently possible , however , because a database application may not have retention functionality . Therefore deletion may not be done .
Is it appropriate ? The question is whether someone would get in trouble for not doing it . It ’ s very rare but it does occur that we have laws that specify that certain types of records must be discarded after a particular period of time .
JB : What would be a good approach to managing the structured data ?
BS : First we would need to know what are these records . A description of business purposes and an understanding of applicable laws and regulations . Then we would have to understand the business use and operational requirements . Do the records have historical value ? So , you set the retention period first and then think about how to implement it . If the implementation requires deletion , you have to look at the database to see if it is capable in some way . Some databases allow you to transfer older transaction records into a history file . The problems come with databases that don ’ t have that functionality and no other retention functionality is built in . In those cases , you have to assess the risk of not deleting against the cost of building in the required functionality . Determine whether the cost outweighs the risks .
JB : Based on interactions with others in my field , they too talk about creating archives tables so you extract the data that needs to be deleted . One of the factors is to what extent how will deletion impact other data .
BS : Yes , that is one of the problems , especially with relational databases . This is why people need to consult with IT . The problem is that it ’ s never going to be on the “ front burner ” for them unless the database is huge and is impacting performance in some way . The traditional motive for paper records — cost of storage — typically does not apply to digital information .
13