Tag Archives: Microsoft

The tale of two (Microsoft) PKIs

Leave a reply

As you know I used to work at Microsoft on areas surrounding cryptography, certificates, protocols and other such things.

While I can’t talk about the internal workings of things I can explain the high level roles the two sets of PKIs Microsoft maintains.

The first PKI is the Microsoft “Product Roots”; these are the keys and certificates that are managed with the sole purpose of being used for product scenarios. You can see the names and serial numbers of these CAs here.

The next is the “IT Root”, that is the one folks see associated with a S/MIME email from a Microsoft employee, some of the corporate websites have SSL certificates from this hierarchy and if you were to inspect the badges of an employee you would see a certificate from here that is used for smart card login and other related scenarios.

The “Product Roots” are trusted only by Windows and applications that have either built on the Windows CryptoAPI or simply imported the roots from the Microsoft Root Program.

The “IT Root” is trusted by third-party browsers because it has been cross signed with a public CA, currently that CA is Verizon the certificate is labeled “GTE CyberTrust”.

I hope that helps clarify for people,

Ryan

MSRC 2718704 and Terminal Services License Certificates

1 Reply

When looking at the certificates the Flame authors used to sign the malware in their Windows Update module one has to wonder what was different between a normal terminal services license certificate and the one they used.

Well getting at one of these certificates is a little awkward as they are kept in the registry vs. in the traditional certificate stores, thank goodness with a little due diligence you can find one.

   
   

 

So from this what can we learn? There are a few things:

  1. Every subordinate CA has the same name “Microsoft LSRA PA”
  2. Every terminal services license has the same name “Terminal Services LS”
  3. The license appears to have a random serial number (even the CA certs did not!)
  4. The license has a 512bit RSA key.
  5. The license is signed with MD5.
  6. No restrictions have been explicitly placed on this certificate.
  7. This license is good for 257 days.
  8. The CRL url is invalid.
  9. The Certificate Issuer url is invalid.
  10. It contains a proprietary extension (1.3.6.1.311.18)

As far as best practices go there are clearly a lot of problems here:

  1. Same name for every CA – the point of a CA is to identify an entity, if every CA has the same name this can’t really happen.
  2. Same name for every subject – while the certificates here are for authenticating entitlement to make a connection and not to authenticate an identity the subject should still be unique.
  3. Over entitled– The certificate contains no Extended Key Usage or other restrictions, as a result this certificate is only as restricted as its issuers issuer is, namely it is good for:
    • Code Signing (1.3.6.1.5.5.7.3.3)
    • License Server Verification (1.3.6.1.4.1.311.10.6.2)
  4. Weak keys – RSA 512 bit was factored in 1999, there should be no CA issuing 512 bit keys period especially in a PKI as powerful as this, especially for one that is good for code signing.
  5. Weak hash algorithm – MD5 has been known to be weak since 1993 and has had increasingly effective attacks published against it ever since — There should be no CA issuing MD5 signed certificates in 2012.
  6. Invalid URLs – This indicates that the CA was not actively managed, it also included internal only file references in the certificate also a no-no and a indicator of the same.
  7. Poorly thought out proprietary extension – The extension is not well thought out, the OID (1.3.6.1.311.18) is an arc for Hydra (Terminal Services) and not a unique identifier (or shouldn’t be at least) for a specific extension. This isn’t harmful but a signal the design didn’t go through much review.

The other items are odd, like the 257 day validity window and the proprietary extension but they are not problematic – at least on the surface.

Well that’s what we know so far.

 

MSRC 2718704 and the Terminal Services Licensing Protocol

3 Replies

There has been a ton of chatter on the internet the last few days about this MSRC incident; it is an example of so many things gone wrong it’s just not funny anymore.

In this post I wanted to document the aspects of this issue as it relates to the Terminal Services Licensing protocol.

Let’s start with the background story; Microsoft introduced a licensing mechanism for Terminal Services that they used to enforce their Client Access Licenses (CAL) for the product. The protocol used in this mechanism is defined in [MS-RDPELE].

A quick review of this document shows the model goes something like this:

  1. Microsoft operates a X.509 Certificate Authority (CA).
  2. Enterprise customers who use Terminal Services also operate a CA; the terminal services team however calls it a “License server”.
  3. Every Enterprise “License server” is made a subordinate of the Microsoft Issuing CA.
  4. The license server issues “Client Access Licenses (CALs)” or aka X.509 certificates.

By itself this isn’t actually a bad design, short(ish) lived certificates are used so the enterprise has to stay current on its licenses, it uses a bunch of existing technology so that it can be quickly developed and its performance characteristics are known.

The devil however is in the details; let’s list a few of those details:

  1. The Issuing CA is made a subordinate of one of the “Microsoft Root CAs”.
  2. The Policy  CA  (“Microsoft Enforced Licensing Intermediate PCA”) on the Microsoft side was issued “Dec 10 01:55:35 2009 GMT”
  3. All of the certificates in the chain were signed using RSAwithMD5.
  4. The CAs were likely a standard Microsoft Certificate Authority (The certificate was issued with a Template Name of “SubCA”)
  5. The CAs were not using random serial numbers (its serial number of one was 7038)
  6. The certificates the license server issued were good for :
    • License Server Verification (1.3.6.1.4.1.311.10.6.2)
    • Key Pack Licenses (1.3.6.1.4.1.311.10.6.1)
    • Code Signing (1.3.6.1.5.5.7.3.3)
  7. There were no name constraints in the “License server” certificate restricting the names that it could make assertions about.
  8. The URLs referred to in the certificates were invalid (URLs to CRLs and issuer certificates) and may have been for a long time (looks like they may not have been published since 2009).

OK, let’s explore #1 for a moment; they needed a Root CA they controlled to ensure they could prevent people from minting their own licenses. They probably decided to re-use the product Root Certificate Authority so that they could re-use the CertVerifiyCertificateChainPolicy and its CERT_CHAIN_POLICY_MICROSOFT_ROOT policy. No one likes maintaining an array of hashes that you have to keep up to date on all the machines doing the verification. This decision also meant that they could have those who operated the existing CAs manage this new one for them.

This is was also the largest mistake they made, the decision to operationally have this CA chained under the production product roots made this system an attack vector for the bad guys. It made it more valuable than any third-party CA.

They made this problem worse by having every license server be a subordinate of this CA; these license servers were storing their keys in software. The only thing protecting them is DPAPI and that’s not much protection.

They made the problem even worse by making each of those CAs capable of issuing code signing certificates.

That’s right, every single enterprise user of Microsoft Terminal Services on the planet had a CA key and certificate that could issue as many code signing certificates they wanted and for any name they wanted – yes even the Microsoft name!

It doesn’t end there though, the certificates were signed using RSAwithMD5 the problem is that MD5 has been known to be susceptible to collision attacks for a long time and they were found to be practical by December 2008 – before the CA was made.

Also these certificates were likely issued using the Microsoft Server 2003 Certificate Authority (remember this was 2009) as I believe that the Server 2008 CA used random serial numbers by default (I am not positive though – it’s been a while).

It also seems based on the file names in the certificates AIAs that the CRLs had not been updated since the CA went live and maybe were never even published, this signals that this was not being managed at all since then.

Then there is the fact that even though Name Constraints was both supported by the Windows platform at the time this solution was created they chose not to include them in these certificates – doing so would have further restricted the value to the attacker.

And finally there is the fact that MD5 has been prohibited from use in Root Programs for some time, the Microsoft Root Program required third-party CAs to not to have valid certificates that used MD5 after January 15, 2009. The CA/Browser Forum has been a bit more lenient in the Baseline Requirements by allowing their use until December 2010. Now it’s fair to say the Microsoft Roots are not “members” of the root program but one would think they would have met the same requirements.

So how did all of this happen, I don’t know honestly.

The Security Development Lifecycle (SDL) used at Microsoft was in full force in 2008 when this appears to have been put together; it should have been caught via that process, it should have been caught by a number of other checks and balances also.

It wasn’t though and here we are.

P.S. Thanks to Brad Hill, Marsh Ray and Nico Williams for chatting about this today 🙂

Additional Resources

Microsoft Update and The Nightmare Scenario

Microsoft Certificate Was Used to Sign “Flame” Malware

How old is Flame

Security Advisory 2718704: Update to Phased Mitigation Strategy

Leaving Microsoft, My Goodbye Letter

Leave a reply

Here is my goodbye letter to all of the amazing people I worked with over the last decade:

It was December 2001 when I came back to Microsoft. I joined the team chartered to build security technologies into Windows; I could think of nowhere else I wanted to be. After all, what other technology company in the world had the opportunity to positively impact the security of so many?

In my time here I have had the honor of working with some of the best and brightest our industry has to offer, working on some of the largest and most ambitious software engineering challenges in the world.

I have had the opportunity to work on platforms for cryptography, public key infrastructure, smart cards, biometrics, network authentication and policy, network isolation, cloud authentication, document signing, code signing, secure boot, volume encryption, enthusiast user experience, helped secure the advertising platform and so much more.

All the while I had the honor (and responsibility) of representing Microsoft in standards forums, working closely with industry partners and leaders to deliver the technology that has laid the groundwork for the consumerization of IT we are experiencing today.

My time here has taught me more than I ever thought it would; as much as the experiences themselves made me better, my greatest lessons came from you. Sometimes these lessons were a result of the folks I worked with respectfully helping me grow, but in many cases they came from simply watching how easy you all make the stuff we do look.

For these lessons I want to thank you.

Ten years later, a new set of challenges are emerging; Certificate Authorities are being forced to re-evaluate how they do business as a result of Advanced Persistent Threats and emerging technologies changing the way trust is communicated on the Internet. These challenges, of course, also represent an opportunity.

As such, I have accepted a position with GlobalSign as their Chief Technology Officer, where I have an opportunity to re-think what it means to be a trusted third-party. My last day will be January 20th.

Please keep up the good work and don’t be a stranger,

– Ryan Hurst
rmh (at) unmitigatedrisk.com
http://unmitigatedrisk.com/
@rmhrisk on Twitter

How to clear the CryptNet cache in Windows 7

8 Replies

OK, so this is going to be geeky and I wouldn’t normally post stuff like this to my Facebook page but for various reasons I can’t post to my blog right now and I want to capture this somewhere.

So in Windows there are several services related to the cryptography, certificates and smartcards; services are able to perform actions for the user and system in the background and enable application developers to do things in a least-privileged way.

One of the core services in these scenarios is the “Cryptographic Services” service; it does a bunch of things including the wire retrievals for CryptoAPI.

Specifically it is the worker for CryptRetrieveObjectByUrl which is used by Windows and other applications to gather evidence necessary to validate certificates, such evidence includes intermediate certificates, CRLs, OCSP responses and a file called commonly referred to as the Windows Certificate Trust List.

This API (at least in Windows 7) maintains a single cache for the whole system of the objects it has downloaded.

These files are kept in a hidden system folder called CryptNetUrlCache, in some cases you may want to test a scenario without relying on the cache, to do that you must flush the cache. The easiest way to do that is to open an administrative command prompt and run the following commands:

cd %WINDIR%\ServiceProfiles\LocalService\AppData\LocalLow\Microsoft\CryptNetUrlCache

attrib .\Content\*.* -s

del .\Content\*.*

attrib .\MetaData\*.* -s

del .\MetaData\*.*

 

%WINDIR%\SysWOW64\config\systemprofile\AppData\LocalLow\Microsoft\CryptnetUrlCache

attrib .\Content\*.* -s

del .\Content\*.*

attrib .\MetaData\*.* -s

del .\MetaData\*.*
%WINDIR%\System32\config\systemprofile\AppData\LocalLow\Microsoft\CryptnetUrlCache

attrib .\Content\*.* -s

del .\Content\*.*

attrib .\MetaData\*.* -s

del .\MetaData\*.*
%WINDIR%\ServiceProfiles\NetworkService\AppData\LocalLow\Microsoft\CryptnetUrlCache

attrib .\Content\*.* -s

del .\Content\*.*

attrib .\MetaData\*.* -s

del .\MetaData\*.*

 

Alternatively you can call this command:

certutil -URLcache * delete

 

No reboot is necessary, next time a component calls the CryptRetrieveObjectByUrl API it will not be able to satisfy that request with the cached data and will be forced to go on the wire.

One of the functions the service offers is the Automatic Update of the root store, a way to validate the cache is not being used is to:

  1. Remove all “Trusted Third Party CertificateAuthorities” from the Computer Account’s store using the Certificate Managementconsole.
  2. Clear the cache as described above
  3. Visit https://www.godaddy.com
  4. in IE
  5. Open Even Viewer\Application
  6. Sort on “Event ID”, find the 4097

Since every time a root is added a new event log entry is created you will see something that says “Successful auto update of third-party root certificate” in the event log, you will also see a few files in the above directories you previously cleared.

This all tells you that new wire retrieval took place and that the cache was not used.

You can of course also use tools like Reg/FileMon as well as Network Monitors to infer much of the same.

 

Hope this helps someone someday,

Ryan

ASKING THE DRUNK IF THEY ARE DRUNK

Leave a reply

For a over decade there has been a slow move towards building networks where only authenticated users can get on (wireless networks are a great example of this), this later moved towards building networks where onlyauthenticated users on managed machines get on, the most recent change to this trend has been to add checks that also require these machines meet a base-line configuration that gives the network administrators some idea of what kind of risks they are exposed to by letting these machines on their networks.

This all makes sense, after all machines that do not have up-to-date antivirus or the latest patches certainly represent more risk to the network than those that do.  A great example of this is actually SQL Slammer, even today five years after the patch for this vulnerability it is still negatively effecting customer environments.

Analysts have given solutions in this space the label of Network Admissions Control (NAC), these systems rely on two basic models one where the behavior of the host is externally monitored and its access is limited if it behaviors out of norm with the networks policy and another where the host makes claims about its configuration that is then evaluated to decide of the host is conformant with the networks policy to see if access should be limited.

Both of these models are examples of what I call “asking the drunk if he is drunk”,  I say this because in both cases the client can control if it displays a behavior (I used to work with a security consultant who would always say “give me your rules and I will comply with them”, this is a example of just that – warning pdf link) or produces a claim that will result in access being given or restricted. The core difference between the two approaches is how rich is the information that is available to make policy decisions on.

The problem with this is if we are to deploy solutions like this we must also accept that it is possible for malware (like root kits or other malicious host software) to change the behavior of the host so that it isn’t “caught” by these checks.

If thats really the case do these systems still have value? I would argue they do, you see they give systems administrators get something they have never really had before, insight into their networks risk profile (how many people on my network do not have this patch, are not running antivirus or have out of date signatures, etc.).

There are other values too, in general users actually “want to do the right thing™” but they don’t necessarily know what that means and these systems help them do just that, in addition these systems often implement concepts of “automatic remediation” so that you can remove the obligation of the user to do anything and instead you just fix it for the user without their interaction.

These reasons underscore why even in computing it’s important to “pull the drunk over and ask him to walk the line”, but does this mean we should accept the status quo? I would say no, although it will never be possible for a running system to evaluate its own state in without potentially being fooled there are lots of things we can do to make the host itself more trustworthy.

As a example many of the technologies in VISTA do just that (and I thinkSymantec would agree), some of the technologies in VISTA that help here today include:

  • Bitlocker – Protects the host from offline attacks of trusted software components and cryptographic keys.
  • Secure Startup – Increases the trustworthiness of the OS by checking to see that core components are authentic as part of the startup sequence.
  • Code Integrity – Increases the trustworthiness of the running state of the OS by checking to see if they are authentic before being loaded.
  • UAC – Reduces the negative impact that can be caused by careless users and/or malware by giving these processes a restricted token.
  • ASLR – Makes it more difficult for malware to attack running code by randomizing where code is running in memory.
  • DEP – Makes it more difficult for malicious code to get and keep running.

This represents just the beginning through, future versions of these conditional network access solutions will also leverage hardware that runs before the operating system (TPM) to verify that the components that implement the above are not tampered with before the operating system even is invoked.

Later systems will even be able to use virtual machines to add even more protection from these sorts of attacks, again that’s not to say that when we get here we will have made these systems fool proof; it’s just not possible(my friend Cem always has great analogies, one of my favorites is “Security is often a exercise of re-arranging deck chairs on a sinking ship” this problem is a example of just that) but that doesn’t mean there is not value in raising the bar and I think the industry is well on its way to do just that.

As a side note, we have deployed the Microsoft solution in this space (Network Access Protection) to over 75,000 hosts over the last few months pretty cool if you ask me.