Explanation of the DBMonitor project

[wibeasley]

I just wrote this email to the first users. I'm posting it here because I think it's a good start to describe the purpose of this tool to people who's databases we might use it for. @bard1536 & @thomasnwilson, feel free to make comments/questions.

I’m adding the four of you to a REDCap project, so you’ll probably get a funny email in the next few minutes.

I’m not moving any of your database from SQL Server to REDCap. The four things REDCap will be doing are:

1. Storing row counts (not the actual data) of each table in your database. The database will be automatically & periodically polled, and the row counts will be deleted. This longitudinal record will help detect if (a) a large batch of records was deleted, or (b) not as many records were added as you expected.

2. Storing the location of where to save the backups. As I’ve discussed with some of you, I want a backup system that’s independent of Peds IS. These backups will be stored as (compressed) CSV files. Anytime you like, you can open these CSVs (with something like Excel) and verify the backups exist. This strategy won’t work for huge databases, but I don’t expect that your specific one will cause any strains.

3. Providing a GUI where you can initiate either step above. The new software will still monitor the record counts and backup to CSV daily. But if there are times when you want run an explicit backup, you should be able to log into REDCap and press a button to initiate it. This feature won’t be complete for another few weeks.

4. Providing a place to see the longitudinal reports of the row counts and the success/failure of the backups.