The Perfect Storm

Its time to get your raincoats and lifeboats - the perfect storm is finished brewing - it is about to rain down upon us.

This may sound dramatic but I think that I may not be conveying the amount of pain that Information Security is about to receive. We will certainly have to step up our game.

Symantec and Verizon have done some interesting research into the underground hacker community and their findings are rather interesting. A bit scary too.

There is an entire community of totally different players that all work together to get from the point where a nerdy kid finds a vulnerability to where a hacker uses that to get into a PC, steal personal information and credit card details, sell them or use them and move on.

So far, it seems, that the community has been quite lazy and have just discarded company information to get to the credit card information and personal information (ID numbers, social security numbers, addresses etc).

This has provided us in Information Security with a perfect opportunity. We have been able to observe how hackers work while they have been taking information that is not our own. Companies that have credit card information have been the ones that were most under attack but those that don't handle credit card information have largely been ignored by hackers except for some members of staff who have been caught out but then they have only lost their own personal information.

There just really isn't a (black/underground) market for information that is not credit card or personal finance related.

However, it was always my feeling that the credit card/personal finance market would become saturated at some stage and the loosely-bound-but-still-very-organised-and-co-ordinated underground market would start to look elsewhere.

Essentially, the infrastructure is there for wide-scale information theft but the will wasn't there. I have thought this for a while my question was always - when will the will be there? When will Jack-the-hacker decide that credit card theft is no longer worth his time and start to deal in company information ?

Adrian Lane from Securosis thinks that the falling prices in the underground economy is humorous. I disagree. I look at it as very scary and the final puzzle-piece.

I think that the perfect storm is about to be unleashed.

Virtualisation - Welcome Back to the 90s.

I've been thinking about this for a while but this blog post by Pascal Meunier pretty much sums up my feelings about Virtualisation.

Back in the 90s when the Internet was new-ish and just becoming important all the machines running it were Unix boxes. (Maybe not all, but most). And a 386 would typically run DNS, sendmail, telnet (shell accounts), ftp and apache. All on the same box.

Security wasn't so tight in those days but it was usually good enough and the box could happily do what it needed to do.

Along came Microsoft and produced the idea of "one box - one service". You can't seriously consider running your domain controller as a file server. What are you thinking? And to put mail on the same box? No way. In fact, your SQL server is running under significant load, chain a few together.

And companies would buy into this concept. Microsoft were happy - more licenses. All the PC guys were happy too - more money. More complexity - more jobs.

Essentially what has happened now is that Moores Law has kicked in and has caught up with the complexity of Microsoft's software to the point where one server box can run multiple applications on it. Imagine that. But Microsoft has planted the one-service-one-box concept so well that it is now part of IT law. File server and mail server on one box? But wait...whats this button over here....? Vir-vir-virtualisation.

And now we have the tools to allow us to once again run multiple applications on one server without having to admit that one-application-one-server never made sense.

To be fair - Virtualisation does have other advantages - running multiple Operating Systems for example, being able to easily move a virtual machine from one box to another (without configuration issues), being able to make a snapshot backup of a system.

But running multiple applications on one box is not a huge win.

Andy sees the light

As per usual the man-in-the-trenches Andy-It-Guy comes up with some excellent observations.

He has found an example of what Bruce Shneier calls movie plot security. What is also known as "whack-a-mole" security or knee-jerk reaction. Essentially, something goes wrong and we put in controls in case it happens again. Then something else goes wrong ... we put in something different. Ad infinitum.

(The name "whack a mole" comes from the game where you have a mallet and you keep whacking plastic moles on the head. Every time you are successful a new mole pops up.)

This is a solution but its not the best solution. And in case you think that it is a terrible solution think about what anti-virus, anti-spam, anti-spyware, firewalls, IPS, patch management etc etc are... point solutions to single issues. Whack-a-mole solutions. Knee-jerk reactions to problems that crop up.

The one technology in the above list that is unfairly listed is Firewall. Best practices state that you should block everything and enable only what you need. But in the past Firewalls were generally configured to block only what was bad and to open up everything else. Then they started to "by-default" block everything in and allow everything out. We have learned our lesson with Firewalls.

Antivirus too is starting to move from a "detect and delete the following..." to a "detect strange happenings from all software... but ignore this that we know is supposed to have extra access."

Note the move from "allow all and block specific known bad" to "block all and allow specific known good".

I think that the challenge going forward will be for us to create an environment where it is possible to tie down exactly what every single person in the organisation does. Make sure that the technology supports this. Make sure that deviations are blocked.

And on top of that allow for agility.

This is not impossible but it won't be easy. But there won't be turn-key technology solutions to be able to achieve this.