Category Archives: Security

Cloud VXLANs and segmentation

The “invisible infrastructure” of VMware cloud is a vision that emphasizes freedom from boundaries

The Virtual Distributed Switch abstracts the data center fabric and provides a sea of ports. vCloud Director (VCD) creates an Org Virtual Data Center (VDC), including allocating compute and storage resources. Tenants/orgs can now provision their own logical network to connect these resources. VCD delegates networking/security control to the vShield Manager, which in turn creates a VDS port group backed by a VXLAN, maps the tenant id to the VXLAN segment id, and connects org VMs to the respective ports in the port group. Additionally, vShield Edge provides multicast services, and maps tenant broadcasts into provider multicasts (using PIM). We now have VXLAN backed logical networks, which are elastic (add/delete vNics/ports on an as-needed basis).

With networking constraints out of the way, VDCs can now span cluster, pod and subnet boundaries, removing one of the major limitations in the data center.

I covered this in my VMworld 2011 sessions on Penetration Testing the Cloud relative to the concerns around segmentation. Risk analysis or threat maps will help refine the topic but in most cases the best security is one that does not impede the ability of the business to operate freely. The ultimate security objective, in other words, should be to create freedom from interference.

While removal of the limitations might sound scary at face value there are many ways to transparently validate segmentation and controls are still effective, even in a multi-tenant environment. I will be presenting on this again several times in the next few months, and then publishing a book early next year with a toolkit and scripts to help.

Save the Library

Once again libraries are under threat of closure. It seems strange that a place of privacy protection and learning could lose support at a time when they are more relevant than ever.

Take the Netflix model of paying a nominal monthly fee in order to check out a movie, for example. Who wants a Netflix account when they could give the same amount to their local library and get far more in return? Libraries do information dissemination without the burden of trying to make a profit for their investors, which has come to mean they don’t have any incentive to track, collect and sell your identity information. They also end up allowing people to share access to a single license but, unlike Netflix, the license is technically owned by the viewers who share access.

Even more interesting is an idea that the notion of a library, as an exchange of information for public access, could be protected by law.

The Act says a local authority which is a library authority must “provide a comprehensive and efficient library service for all persons . . . whose residence or place of work is within the library area of the authority or who are undergoing full-time education within that area”. Its stock of “books and other printed matter, and pictures, gramophone records, films and other materials”, must be “sufficient in number, range and quality to meet the general requirements and any special requirements both of adults and children”.

That sounds like the library is the school. It might seem crazy to try and legislate the quality of information until you read how Isaac Asmiov described the library in a letter to new patrons of one in Troy, Michigan:

Libararies are your best friend

An updated version is on YouTube from Piers Cawley, who wrote and performed a song called “Child of the Library” at OSCON 2011 and then received a standing ovation:

Free exploit automation: Pmcma released

Funny intro in the README

Is this tool for me ?
———————

[…]

As a script kiddie, you may have found a piece of code you don’t understand on the internet, but are nonetheless decided to go to jail.

In all those cases, and surely many others, Pmcma was probably made for you.

I think they mean that if you run Pmcma on code without authorization and get caught you will go to jail. The decision to go to jail? That sounds like a protest. I don’t think that fits with the motive of someone who wants to run scripts in the sense of a “kiddie”. Perhaps it could be translated into French like this:

En tant que pirate adolescent vous voulez tester le logiciel sur Internet et ne se soucient pas d’aller en prison.

Ok, that’s my attempt at Canadian French, but still…I put the emphasis on being unaware of consequences rather than making it a decision to go to jail.

Anyway, Pmcma offers to automatically write exploits for flaws it finds in software (given it has root privilege) without the need for sourcecode.

Restaurants That Stalk Online Commenters

Interesting quote from the owner of a San Francisco restaurant.

Weinberg says in her blog that: “With a bazillion places online to tell us how badly we sucked, we do take it very personally”. “We scour the sites, cyber-stalking our customers.” She isn’t joking about the cyber-stalking.

When they see a negative comment, Weinberg and her team will track the customer through cyber-space to see what other restaurants they frequent and how they have rated them, before determining whether the complaint should be taken seriously. If they get the feeling that something should change, they change it. “Both online comments and in-house feedback usually reflect if the menu needs tweaking,” she says.

It sounds like they take the comment seriously because they take the trouble to track the customer. Then they determine whether it is a false positive. What restaurants need like a behavioral index tool. In other words they could save a lot of time if they had a simple reputation engine that gave them a score for an identity based on a list of other restaurants with comments from the same identity. Then they wouldn’t have to take every negative comment seriously, only the ones from identities they “respect”.

Then again this indicates a serious logical fallacy as a filter. It begs the question of how they respond to comments from identities they can recognize even without tracking them. Do they think it’s wise to judge the person before they listen to the message?

What if they designed a filter instead to be based on details of an event? When a commenter gives specific feedback about a taste, a detail that only a real patron could know, then they would know to take the comment seriously. A generic comment would be ignored. The flip side of this is that the restaurant would have to accommodate change in their menu and/or service to allow comments to be unique.

If they serve up a hot dish of key management, so to speak, then they can easily track the day and time the customer ate, and they can focus on the facts of the comment rather than the person writing a comment. A win-win; valuable feedback for restaurants and freedom (from stalkers) for their customers.

And just for reference, here is the restaurant owner’s FAQ, which might give you some insight into what she really thinks when people comment…

Q. Wow, Anna did you notice how big this space is? That’s a ton of seats to fill…

A. Yes a#$%##e I noticed how big it is.

Q: It really doesn’t look like you will be done by September. Or even this year.

A. Yes a###%^^e I noticed we are a little behind.

Q: Isn’t it like, impossible to find this many good staff?

A. Yes a$%$&&hole. It’s very hard to find good staff these days.

Q: Is that where the bar is going?

A. Yes a$$%%@e, that’s where the obviously brand spanking new bar is going. It’s right there in front of you.