So users have a top level thing “project”. They can have many projects.
Underneath projects they have many other types of associated things (30+) and under those types of things are another tier of associated things.
If a user makes extensive use of a project, the result will be around 1000 things related to a project.
Now if they choose to delete that project, I have to recursively search and deleted down through all things.
At a minimum this will cost me 0.1 WU (cost to delete a thing)*1000 =100WU just to keep the DB clean. Then add in searches etc.
Alternatively I can just set a flag on the top level project “deleted” = “yes” and be charged only 0.5WU.
In a sane world where we’re not nickled and dimed for everything I’d of course opt for the deletion but financially I feel as though I’m encouraged not to delete in this case.
Any thoughts? Will this bite me down the line with DB performance etc.?
If budget is a major constraint, use the “deleted” flag approach but implement a scheduled cleanup process via backend workflows to delete old “deleted” projects and their associated data periodically. This gives you the best of both worlds: minimizing immediate costs while maintaining long-term database performance.
Interesting, that was going to be my next question but you answered before I even asked
A strange situation then where regardless of the approach, it’s financially sensible at least to not delete.
Of course the risk of DB performance degrading over time.
I suppose the solution is to wait for a “light” month in terms of WU usage, where you’re still within your plan’s limits and deal with the deletions at the end of that month.
Crazy this is even something that needs any thought from us as the customer though.
