Security is not an all-or-nothing proposition. And that’s part of the problem. It creates blind spots; gaps in vulnerability. Partly because of the inherent complacency that after a company institutes a new security initiative that hackers will be held at bay, or the employees won’t be tempted to make off with a database or a hundred other internal or external threats.
I have long promoted that security is as much about planning and process as it is about the various solutions that are deployed to protect networks, data, and other assets.
Security is no longer a wall. Stick up a firewall and your customer/user data won’t get leaked. Actually it’s more like a sandwich. And good security initiatives are like a good Dagwood sandwich—layered and integrated. But even the best sandwiches have holes (if they didn’t, you’d be eating a brick).
So, just as you wouldn’t eat the bread, then eat the meat, then the cheese, and then dip your fingers in the mustard, you shouldn’t deploy firewalls in parallel with intrusion detection, identity management, access management, etc… The key term here is in “parallel.” As part of your strategic planning, you must recognize that although each has a certain and important function, if they are simply working in parallel and not collaboratively, you are creating significant blind spots.
To illustrate I want to focus on a typical internal threat of unregulated user access. What happens when employees use web access to process corporate data, including on their own devices? This issue is more than just updating sales call results via salesforce.com or logging in to social media to post about an upcoming promotion. It’s about control and maintaining the control of the IT environment.
With so much data shared beyond the traditional firewall, cloud identity management (credentialing and authorization) is a good starting point, but if you simply allow users a means to enter the castle, what they do once inside still creates risk. Strong identity management not only provides the credentialing gatekeeper (user name, passwords), but must apply use roles-based provisioning to further segment users. In earlier blogs I’ve discussed how not all users are created equal and therefore should not be allowed generic universal access. To further mix my metaphors, think of it like a hotel. Only guests are allowed passed a certain point-and each has Cloud Access to their own personal space. But only certain guests are allowed into premium areas such as spas or penthouse rooms. The same is true for users and the assets on your extended network.
It is important to recognize that every single user has a unique relationship to your IT environment. Employees, partners, customers, etc… each have their own agenda and their own needs. Credentialing gets them in the door. Provisioning gets them a proverbial room in the hotel. But as mentioned earlier on, that still leaves blind spots. The next obvious step is to integrate access management (single sign on) that automatically leverages roles based provisioning.
By having a single-sign on portal based off of an Active Directory or LDAP controlled identity manager does two things. It provides access to only the relevant applications that a user might need based on their role. Marketing would not see the MySQL apps, Research would not get ADP. However they might need to access Workday or Concur to help manage their timecards. But their identity limits them to just their responsibility (or if they run a department, their subordinates as well). The second provision is that access management can also control non-SAML based (Security Assertion Markup Language) applications and websites. Not every application is cloud SaaS these days. Upwards of 85% of all applications are non-SAML based (including the legacy application sitting on your server right now!), but still require the same protections. Now you still need SAML, or OAuth or Liberty federation for some applications (especially SaaS that contains sensitive data like GoogleApps or Office365, but to get maximum benefit from integrate IDM/AM you need to incorporate all the apps.
For instance, take WordPress. It requires a user name and password. It’s not federated via SAML. Although you may not be sharing state secrets on your blog site, it still can be hijacked for disreputable purposes, or be used as a gateway to gain entry into the network. Putting it under the protection of single sign on (under the provisioned protection of identity management), the passwords are secure and the access is limited to only those who need to see it on a user-centric portal. And practically every legacy application is the same way (do you still have MSOffice 2007 on each workstation?).
This plays into two camps of benefit. One is based on productivity and accountability; but that’s on the user (or at least the user’s boss!). To the IT administrator it is all about regaining and maintaining control; not just who gets to see what, but when and how (and leaving the necessary information trail to remediate issues or provide compliance reporting). But in terms of covering those blind spots, you cover all the application (or intranets) directly controlled on your servers, all your SaaS and web-based apps, and all relevant websites. You now have the visibility to know Rodney from business development is certainly spending a lot of time on Facebook. But if we take this coverage to next logical level and truly erase blind spots, you should think of unifying IDM/AM with other security initiatives such as real time intrusion detection and event archiving as well. That way the visibility extends to knowing when Rodney is logging on and if it is truly him; whether he is using assets at the office, at home or on the road—or if it is someone spoofing his account from an IP address in Outer Mongolia.
There are obviously a great many benefits of an integrated identity and access management deployment, especially when managed from the cloud (security as a service). The cost, the resources, the expanded enterprise capabilities alone make it quite attractive and highly manageable for a SMB or mid-sized company. The CBS Morning Show just did a story on how smaller companies are now squarely in the sights of hackers because of the lower security barriers. The kind of initiative I have described above is typically based on one or the other enterprise solution being on premise (very expensive!). However, evolution and revolution from the cloud is currently available and ready to leverage and collaborate with any existing security plan. (I should know, right?)
Obviously the cure to removing blind spots is better visibility. Better visibility can only be achieved when you can see the entire playing field in front of you. And that can only be achieved when you security solutions work together. And now that is a fast and attainable achievement from the cloud.
Who is now craving a Dagwood sandwich