I have created a dashboard (with basic auth), which contains some potentially sensitive information. I ended up using a desktop on my local network to act as a “server” given the lack of infrastructure in-house. So if I give my management the IP address and port of that machine, they can log-in via the preset username and password I send via encrypted email. The limitation is they need to be logged onto our network obviously.
This solution would be an interim solution for once we procure our actual solution. So we will probably use this for a year, maybe two.
I load the datasets in Python on that machine before initiating the dashboard, instead of loading the datasets within the Callback Inputs of the dash app itself. I also used the werkzeug in the
read.csv command. Not sure if any of these actually help anything.
Is there anything I should be doing to make it more secure? Would it make sense to combine this with Flash-login (if it’s even possible?) We have Edge, Explorer and Chrome, and each one says “Not secure” in the top (for good reason I imagine).