Research

I’m happy to report that we have finished the process description posts for the Malware Analysis Quant project. Not all of you follow our Heavy Feed (even though you should), so here is a list of all the posts. The Malware Analysis Quant project addresses how organizations confirm, analyze and then address malware infections. This is important because today’s anti-malware defenses basically don’t work (hard to argue), and as a result way too much malware makes it through defenses. When you get an infection you start a process to figure out what happened. First you need to figure out what the attack is, how it works, how to stop or work around it, and how far it has spread within your organization. That’s all before you can even think about fixing it. So let’s jump in with both feet. Process Map Confirm Infection Subprocess This process typically starts when the help desk gets a call. How can they confirm a device has been infected? Notification: The process can start in a number of ways, including a help desk call, an alert from a third party (such as a payment processor or law enforcement), or an alert from an endpoint suite. However it starts, you need to figure out whether it’s a real issue. Quarantine: The initial goal is to contain the damage, so the first step is typically to remove the device from the network to prevent it from replicating or pivoting. Triage: With the device off the net, now you have a chance to figure out how sick it is. This involves all sorts of quick and dirty analysis to figure out whether it’s a serious problem – exactly what it is can wait. Confirm: At this point you should have enough information to know whether the device is infected and by what. Now you have to decide what to do next. Confirm Infection Process Descriptions Based on what you found you will either: 1) stop the process (if the device isn’t infected), 2) analyze the malware (if you have no idea what it is), or 3) assess malware proliferation (if you know what it is and have a profile). Analyze Malware Subprocess By now you know there is an infection, but you don’t know what it is. Is it just an annoyance, or is it stealing key data and presenting a clear and present danger to the organization? Here are some typical malware analysis steps for building a detailed profile. Build Testbed: It’s rarely a good idea to analyze malware on production devices connected to production networks. So your first step is to build a testbed to analyze what you found. This tends to be a one-time effort, but you’ll always be adding to the testbed based on the evolution of your attack surface. Static Analysis: The first actual analysis step is static analysis of the malware file to identify things like packers, compile dates, and functions used by the program. Dynamic Analysis: There are three aspects of what we call Dynamic Analysis: device analysis, network analysis, and proliferation analysis. To dig a layer deeper, first we look at the impact of the malware on the specific device, dynamically analyzing the program to figure out what it actually does. Here you are seeking perspective on the memory, configuration, persistence, new executables, etc. involved in execution of the program. This is done by running the malware in a sandbox. After understanding what the malware does to the device you can start to figure out the communications paths it uses. You know, isolating things like command and control traffic, DNS tactics, exfiltration paths, network traffic patterns, and other clues to identify the attack. The Malware Profile: Finally we need to document what we learned during our malware analysis, packaged in what we call a Malware Profile. With a malware profile in our hot little hands we need to figure out how widely it spread. That’s the next process. Malware Proliferation Subprocess Now you know what the malware does, you need to figure out whether it’s spreading, and how much. This involves: Define Rules: Take your malware profile and turn it into something you can search on with the tools at your disposal. This might involve configuring vulnerability scan attributes, IDS/IPS rules, asset management queries, etc. Define Rules: Process Description Find Infected Devices: Then take your rules and use them to try to find badness in your environment. This typically entails two separate functions: first run a vulnerability and/or configuration scan on all devices, then search logs for indicators defined in the Malware Profile. If you find matching files or configuration settings, you need to be alerted of another compromised device. Then search the logs, as malware may be able to hide itself from a traditional vulnerability scan but might not be able to hide its presence from log files. Of course this assumes you are actually externalizing device logs. Likewise, you may be able to pinpoint specific traffic patterns that indicate compromised devices, so look through your network traffic logs, which might include flow records or even full packet capture streams. Find Infected Devices: Process Description Remediate: Finally you need to figure out whether you are going to remediate the malware, and if so, how. Can your endpoint agent clean it? Do you have to reimage? Obviously there is significant cost impact to clean up, which must be weighed against the likelihood of reinfection. Remediate: Process Description Monitor for Reinfection One of the biggest issues in the fight against malware is reinfection. It’s not like these are static attacks you are dealing with. Malware changes constantly – especially targeted malware. Additionally, some of your users might make the same mistake and become infected with the same attack. Right, oh joy, but it happens – a lot. So making sure you update the malware profile as needed, and continuously check for new infections, are key parts of the process as well. Monitor for Reinfection: Process Description At this point we’re ready to start Phase 2 of Quant, which is to take each of the process steps and define a set of metrics to

A flaw in the Oracle database has been disclosed, whereby the Oracle System Change Number (SCN) – a feature that helps synchronize database events – outgrows its defined limits. The SCN is an ever-increasing sequence number used to determine the ‘age’ of data. It is incremented automatically by 16k per second to provide a time reference, and again each time data is ‘committed’ (written to disk). This enables transactions to be referenced to the second, and ordered within each second. As you might imagine, this is a very large number, with a maximum value and a maximum increase per day. If the SCN passes its maximum value the database completely stops. The new discovery concerns the SCN. I’ll get more into the scope of the problem in a second, but first some important background. When I started learning about database internals – how they were architected and the design of core services – data integrity was the number one design goal. Period! Performance, efficiency, and query execution paths were important, but actually getting the right data back from your queries was the essential requirement. That concept seems antiquated today, but storing and then retrieving correct data from a relational system was not a certainty in the beginning. Power outages, improper thread handling, locking, and transactional sequencing issues have all resulted in database corruption. We got transactions processed in the wrong order, calculations on stale data, and transactions simply lost. This resulted in nightmares for DBAs who had to determine what went wrong and reconstruct the database. If this hits an accounting system suddenly nothing adds up in the general ledger and the entire company is in a panic at the end of the quarter. We can normally take data consistency for granted today, thanks to all the work that went into relational database design and solving those reliability problems in the early years. One of the basic tools embedded into relational platforms to solve data consistency issues is the sequence generator. It’s an engine that generates a sequence of numbers used to order and arrange events. Sequence numbers provide a mechanism for synchronization, and help provide data consistency within a single database and across many databases. Oracle created the SCN many years ago for this purpose, and it’s literally a core capability, upon which many critical database functions rely. As an example, every database read operation – looking at stored data – compares the current SCN with the SCN of the data stored on disk to ensure data was not changed by another process during the query. This ensures that each operation in a multi-threaded database reads accurate data. The SCN plays a roll in the consistency checks when databases are brought online and is core to database recovery in the event of corruption. In a nutshell, every data block in a database is tied to the SCN! Now back to the bug: This flaw was discovered as a result of a backup and recovery feature abnormally advancing the SCN by a few billion or even a few trillion. For most firms this will never be an issue, as the number is simply too large for a few extra billion to matter. But for large organizations who have designed their databases to synchronize using common SCNs the possibility of failure is real – and the impact would be catastrophic. At this time Oracle has both patched the flaw during recovery where the number is erroneously advanced and changed the database to double the SCN range. Just as importantly, the provided the patch quickly. The patch appears to fix the bug and with the increased SCN range we assume this problem will never occur in a normal setting. The odds are infinitesimally small. What has people worried is that attackers could leverage this into a denial of service attack and disable a database – or possibly every linked database in a cluster – for an extended period. There are a couple known ways to exploit the vulnerability so patch your systems as soon as possible. What worries me even more is, with this focus on the SCN, that researchers might discover new ways to attack inter-database SCN synchronization and corrupt data. It’s purely speculative on my part, but this capability was designed before developers worried much about security, so I would not be surprised if we see an exploit in the coming months. A couple closing comments: The InfoWorld article that broke news of this flaw is excellent. It’s lengthy but thorough, so I encourage you to read it. Second, if your environment relies on inter-database SCN you need to do two things: level set security across all participating databases, and start looking at a migration plan to reduce or eliminate the inter-database dependency to mitigate risk. For most firms I know that rely on the SCN, the best bet will be to tighten security, as the rewrite costs to leverage another synchronization method would be prohibitive. Finally, Oracle assigned a risk score of 5.5 to CVE-2012-0082. Does that sound accurate to you? Once again Oracle’s risk scores do a poor job of describing risk to your systems, so take a closer look at your exposure and decide for yourself. Share:

I think I need to ban Mike from Arizona. Scratch that – from a hundred mile radius of me. A couple weeks ago he was in town so we could do our 2012 Securosis strategic planning. He rotates between my screaming kids and Adrian’s pack ‘o dogs, and this was my turn to host. We woke up on time the next morning, hopped in my car, and headed out to meet Adrian for breakfast and planning. About halfway there the car sputtered a bit and I lost power. It seemed to recover, but not for long. I popped it into neutral and was able to rev up, but as soon as there was any load we stalled out. I turned around and started creeping toward my local mechanic when it died for good. In a left turn lane. A couple workers (they had a truck but I couldn’t see what tools they had to identify their work) offered to help push us out of the road. Seemed like a good idea, although I was arranging our tow at the same time. I kicked Mike out, hopped in the driver’s seat, and was waiting for a gap in traffic. They weren’t. These dudes were motivated to get us the hell out of their way. Here I am on the phone with the tow company, watching Mike’s face as he decided the rest of us were about to get creamed by the traffic speeding our way… with him on outside the car. I was wearing my seatbelt. We made it, the tow truck showed up on time, and I quickly learned it was what I expected – a blown fuel pump. My 1995 Ford Explorer was the first car I ever bought almost new (a year old, under 25k miles). I had it for about 16 years and it showed it. Living in Colorado and working with Rocky Mountain Rescue, it drove through all sorts of off-road conditions and on rescue missions (including roads closed due to avalanche quirks) that would have pissed off my insurance company. Anyway, despite my emotional attachment, the repair costs were over my mental limit, and it was time to find a younger model. I briefly toyed with minivans but just couldn’t do it. Logically they are awesome. But… err… it’s a friggin’ minivan. I then moved on to SUVs, even though they aren’t nearly as practical. I have rescue deeply ingrained into my brain, and it’s hard for me to not get something with 4WD. And yes, I know I live in Phoenix – it isn’t exactly rational. The GMC Arcadia wasn’t bad. The Dodge Durango drove like my 1980’s Chevy Blazer. The Mazda CX-9 drove well but couldn’t handle our car seat requirements. Eventually I ended up with another Explorer… but damn, they have improved over 16 years! Two words – glass cockpit. Ford is really ahead of most of the other car manufacturers when it comes to telematics. Aside from the big screen in the middle, two others are integrated into the dash to replace analog instruments. They actually issue software updates! Sure, they might be due to the bugs, but late last year I decided I would do my darned best to avoid buying anything with a screen I couldn’t update. Aside from all the cool software stuff, it comes with tons of USB ports, charging ports, and even a built-in 110V inverter and WiFi hotspot so the kids can play head-to-head games. And safety systems? I have… for real… radar in every direction. Blind spot, backup, cross traffic, and even a nifty “you are about to ream the car in front of you up the tailpipe, maybe slow down” alert. It also… er… drives and stuff. Mileage isn’t great but I don’t drive much. And when my phone rings the brakes lock up and the wipers go off, but I’m sure the next software update will take care of that. Almost forgot – the Mike thing? One of the first times he was out here my kid got stomach flu and Mike had to watch her while I took client calls. Then there was the time he had to drive me to the emergency room in DC. Then there was the time we had to end our video session early because I got stomach flu. You get the idea. He’s a bad man. Or at least dangerous. On to the Summary: Webcasts, Podcasts, Outside Writing, and Conferences Rich on How to Monitor Employees Without Being a Perv. I still can’t believe they let me use that title. Mort on counterattacks at CIO Magazine. Mike also quoted at CIO – this time on cloud security. Favorite Securosis Posts We didn’t write much this week, but here’s an old post I’m about to revive. Principles of Information Centric Security. Other Securosis Posts Oracle SCN Flaw. Incite 1/19/2012: My Seat. Censored #sopa. Network-based Malware Detection: The Impact of the Cloud. Favorite Outside Posts Adrian Lane: InfoWorld’s ‘Fundamental Oracle Flaw’ post. Really well done. Mike Rothman: Eating the Security Dog Food. The only way to really lead (sustainably, anyway) is by example. Wendy makes that point here, and it’s something we shouldn’t ever forget. If policies are too hard for us to follow, how well do you expect them to work for users? Project Quant Posts Malware Analysis Quant: Process Descriptions. Malware Analysis Quant: Monitoring for Reinfection. Malware Analysis Quant: Remediate. Malware Analysis Quant: Find Infected Devices. Malware Analysis Quant: Defining Rules. Malware Analysis Quant: The Malware Profile. Malware Analysis Quant: Dynamic Analysis. Malware Analysis Quant: Static Analysis. Malware Analysis Quant: Build Testbed. Research Reports and Presentations Tokenization Guidance Analysis – Jan 2012. Applied Network Security Analysis: Moving from Data to Information. Tokenization Guidance. Security Management 2.0: Time to Replace Your SIEM? Fact-Based Network Security: Metrics and the Pursuit of Prioritization. Tokenization vs. Encryption: Options for Compliance. Security Benchmarking: Going Beyond Metrics. Top News and Posts Symantec Acquires LiveOffice. Norton Source Code Stolen in 2006. Feds Shutdown Megaupload, Bust Founder. Training employees – with phishing! Internet SOPA/PIPA Revolt:

Before we get to the Incite we should probably explain why it’s a day late. Like many other sites we have huge issues with PIPA and SOPA, so we took down our site yesterday in protest. We don’t expect the big companies with big lobbying budgets to give up, so we need to keep the pressure on. Copyright holders have a right to protect their content, but not at the cost of our freedom and liberty. Period. Now back to our regularly scheduled pot stirring. Growing up I spent a lot of time in our den. It was a pretty small room, in a pretty small house, but it’s where the TV was. First it was a cabinet-style tube TV. Remember those? Yeah, you kids today have no appreciation for the TV repair man who showed up to fix your TV with a case full of tubes. Then we got a 15” model with cable, and my brother and I spent a lot of time in that room. The furniture was pretty spartan as well. We had a chair and we had a couch. The couch was, uh, uncomfortable. A plaid model with fabric that was more like sandpaper. Not that microsuede stuff we see today. So amazingly enough, there was a lot of competition for the chair. I usually won. OK. I always won, mostly because I was older and a lot bigger. That may have been the only positive to having a childhood weight problem. I always got the chair. Even if my brother was there first. I’d just sit down. Yes, right on him. It didn’t take long for him to realize I wasn’t moving and my bulk wasn’t going to get any more comfortable. After about the zillionth time I used this approach to get the chair, and my Mom got (justifiably) fed up with my brother crying about it, she instituted a new policy. You could call “my seat” once you entered the den, and you’d have to respect the call. Kind of like calling “shotgun” to sit in the front of the car. My brother may have been small, but he was quick. And more often than not, he beat me to the den and called the seat. I wasn’t happy about it, and when the babysitter was there I’d forget the rule. But inevitably I’d suffer the consequences when Mom got home. So it was funny to see XX2 sit in the passenger side captain’s chair in the van over the weekend. That’s where XX1 usually sits. The little one had this look, like she ate a canary, sitting in that seat. XX1 was not happy at all. I’m not sure whether it was because she likes her seat or that XX2 got the best of her. So the squealing started. And I’m not too tolerant of squealing. For a second I thought about instituting the my seat policy for the van. But that’s overkill for now. The girls don’t physically bully each other, and even if they did, I’m not sure XX2 wouldn’t win her fair share of battles. Though it did give me a chuckle to remember the old days of abusing my little bro. Speaking of which, it’s probably time to take him out to dinner, since I’m still running a huge karma deficit with him. Suffice it to say sitting on him was probably the nicest thing I did when we were growing up. -Mike Photo credits: “I gotta get me one of these!” originally uploaded by sk8mama Heavy Research We have been busy blasting through process descriptions for the Malware Analysis Quant project. Here are the last week’s posts, which zero in on the Malware Proliferation subprocess: Defining Rules Find Infected Devices Remediate Monitoring for Reinfection You can find all the posts on the Project Quant blog. We have also finished up our Network-based Malware Detection series, so here is a link to the last post, on assessing the impact of the cloud. Yes, the forecast is cloudy. Ha ha. Network-based Malware Detection: The Impact of the Cloud As always you can get our Heavy Feed via RSS, where you can access all our content in its unabridged glory. Incite 4 U Revisiting the STRATFOR breach: Looks like Nick re-graded the folks at STRATFOR on their breach response and it went from a B- to a D-. Personally I think that’s too harsh, but ultimately it’s a subjective opinion and Nick is entitled to his. The key is constant communication, which STRATFOR failed at. It seems they spent the two weeks totally rebuilding their infrastructure, as they should have. I also liked the video from their CEO and agree it should have come earlier, if only to initially accept responsibility at the highest level. Then you communicate what you know as you can. I guess everything is relative and I personally think STRATOR did an okay job of response. You can always improve, and you should learn by what they didn’t do well, so you can factor that into your own response plans. – MR Unbreakably irresponsible: I think Adrian is going to cover the latest Oracle security flaw/patch in more detail, but I want to address a long-standing pet peeve I have with the big O. First, let’s give them credit for getting this out relatively quickly, even though it isn’t something that will (probably) affect a large percentage of Oracle customers. Then again, the ones most at risk tend to have 3-4 letter acronym names. Knowing that some flaws are ignored for years, it’s nice to see a relatively quick response – even if it may be due to the press being involved from the beginning. But that isn’t my peeve. You’ll notice that patches only go back a few versions for Oracle 10 and 11, and aren’t available for anything earlier. Oracle reps have told me (not that we talk much anymore) that they don’t believe a significant number of customers are running older versions. And if they are, since said versions are out of support, those customers are