1. Explain the differences between public, private, and community clouds. What are some of the factors to consider when choosing which of the three to use?

2. How do cloud threats differ from traditional threats? Against what threats are cloud services typically more effective than local ones?

3. You are opening an online store in a cloud environment. What are three security controls you might use to protect customers' credit card information? Assume that the information will need to be stored.

4. Define TNO. Name three types of data for which one should want TNO encryption.

5. How do cloud services make DLP more difficult? How can customers wishing to enforce DLP mitigate this issue?

6. You run a website in an IaaS environment. You wake up to discover that your website has been defaced. Assume you are running a web server and an FTP server in this environment and that both an application proxy and a firewall sit between those servers and the Internet. All of your VMs are running SSH servers. What logs might help you determine how the website was defaced? What kind of information would you look for?

7. Sidebar 8-2 shows that personal biographical information-addresses, phone numbers, email addresses, credit card numbers, etc.-can not only be used by attackers to hijack accounts but can also be collected from one hijacked account to help an attacker gain access to the next. How can you protect yourself against this kind of attack? What can cloud providers change to mitigate such attacks?

8. Describe an FIdM authentication system for which you have been a Subject. What organization acted as the IdP? What service acted as the SP?

9. Name three security benefits of FIdM over requiring users to use a new set of credentials.

10. Why is it important to sign SAML Assertions? Why is it not important to sign OAuth Access Tokens?

11. In OAuth, what attack does the Client Secret mitigate? Why do you think the Client Secret is optional for Public Clients?

12. Name four services that might allow you to control a VM in an IaaS environment. What entity controls each service?

13. What are some characteristics of systems in which you would expect application whitelisting to work well? What about systems in which you would expect it to not work well?

Sidebar 8-2 One Man's Single Point of Failure:

In August 2012, journalist Mat Honan had his digital life turned upside down. He was playing with his daughter when his Apple iPhone suddenly shut off. When the phone rebooted, all of his data were gone. Luckily, Honan had set the phone to regularly backup to his Apple laptop, so he wasn't concerned. Soon after he opened the laptop, the screen went gray, and he knew he had a real problem. Before long, Honan discovered that, in addition to his phone, his laptop and Apple iPad had been wiped, and that his Gmail and Twitter accounts had been hacked as well. Here's an abridged version of how it happened [HON12]: The hackers started with the original target, which was Honan's Twitter account ("@mat"). The Twitter account linked to his personal website, which in turn listed his Gmail address. When the hackers went to Gmail to attempt to reset Honan's account password, Gmail showed them Honan's obscured emergency alternate email address:

The hackers correctly guessed that the me.com address would be Honan's username for Apple iCloud, a service that ties all of a user's Apple devices together with data stored at Apple's data centers. Because the hackers had perpetrated attacks like this before, they knew that the only additional information they would need to hack the iCloud account would be Honan's mailing address and the last four digits of his credit card number. The mailing address was easy enough: They just searched the who is record for Honan's website. To get the last four digits of the credit card, the hackers went to Amazon. They correctly assumed that Honan's Amazon username would be his Gmail address and, given that and the information they already had, they were able to trick Amazon into showing them the last four digits of the credit cards associated with the account.

Once the hackers had those four digits, they had all they needed to get Apple customer service to let them into Honan's iCloud account.

Thanks to a very useful iCloud security feature that allows users to remotely wipe Apple devices in the event of theft, the attackers were able to delete all of Honan's data from his devices within minutes.

Perhaps the most amazing part of this story is the way Honan found out how the attack went down: The hackers told him. One of the hackers reached out to him and, in exchange for a promise not to press charges, detailed the whole event. While one can take a number of valuable security lessons from this story, identifying and eliminating single points of failure is an important one. The obvious single point of failure is the linkage between Honan's devices and his iCloud account. He used his Apple laptop to back up his Apple phone, and he allowed his Apple iCloud account permission to remotely wipe both the laptop and phone. But on top of that, it was because all of his accounts were intertwined-albeit in a nonobvious way-that the attack was even possible.