Friday, June 12, 2020

Fresh Rethink

In the technical sense, what is security? Or, what are the most common categories of activity in which the issue comes up? Secure coding practices? Networks? Email? Insider threats? Domain registration/hijacking? All those are contenders but let's look at four examples I've personally seen or heard testimony regarding from those that were involved.

DDOS ATTACKS

Distributed Denial Of Service attacks, in which a bad guy tries to knock you off the air, are persistent and automated. There are many protection services such as Cloudflare that offer access to defense networks for a fee. Your internet service or cloud provider may have filtration devices (network appliances) in a network operations center that can be temporarily effective. Your administrator - if you have one - may know of onboard options in your server or desktop/tablet/phone operating systems that can help, like firewalls, but recall that firewalls are only as good as the rules maintained by the administrator. Also, firewalling is not usually a network-related but a local damage control approach, at least from the end user's perspective.

RANSOMWARE

A stealth attack or a click by an employee on a bad email link can result in the presence of an outside program that can encrypt valuable information. The perpetrator then demands a fee for MAYBE restoring the data (and the encryption method may have inadvertently destroyed some of the data in the process). I think the only way to protect against this sort of thing that offers a high degree of success is an effective backup strategy. Saving all data, configurations and network information in secure, out-of-band or offsite areas gives you the ability to bounce back from most issues. It's business continuity.

CLOUD ISSUES: disappearance

I was told a story by someone involved that a company had a network of virtual computer instances along with virtual routers and switches at a major cloud provider. The company didn't pay its provider bill one month. A few days after the deadline, 32 instances and their related virtual network components simply went away. Vanished. Backups existed and were current. But it took a week to rebuild all the nodes of various kinds. Could your organization survive a week of customer interruption? The cause was an accounting error, producing a sort of security issue. But it's really a continuity issue.

CLOUD ISSUES: hypervisor code

I was witness to a problem at a provider that ran a popular hypervisor on the large servers on which it created virtual computers that were then rented to end users. A coding problem caused by the company that wrote and maintained the hypervisor program was identified that took about ten days to patch. For those ten days we were inundated with frightened questions from customers who were potentially exposed to data exfiltration and we couldn't do a thing for them. If you have sensitive data or trade secrets this may be an argument against cloud structure and for being in control of your own bare metal.

My constant individual user advice: go to the PUB!

PASSWORDS - change them often, with no dictionary words allowed. No storage in the clear (on the network)!
UPDATES - check operating systems, applications, firmware and processor issues regularly.
BACKUPS - check for successful and readable second copies of anything important, and store in some way unconnected to the primary network. Constant connection from primary to backup location could be a problem, allowing an attack to encrypt or erase backups in addition to originals.