Week 5 – Halfway through the last term!

The end of this week marks the downhill side of my final term at Bellevue University. Shortly I will have finished up my Master’s Degree in Cybersecurity. While this is a large accomplishment, my final courses have continued to reinforce that there is always something new to learn. Writing these blog posts is a portion of the assignments for Current Trends in Cybersecurity. The other part of the course deals with threat modeling, something I have little experience with.

Threat Modeling is a concept that requires developers to always study and potentially mitigate threats that their software is susceptible toward. It is an interesting topic and one that has sparked some interest in me. The book and coursework is based on principles established by Microsoft and I have learned a great deal. However, it has also raised several questions for me. The most pressing question is – are startups using threat modeling in their products?

With the startup culture taking over silicon valley and most other development shops across the world, programming practices have changed substantially. Companies now work in sprints, and incubate products quickly to get them out the door. Often times startups refer to their products as minimum viable product (MVP), meaning there are just enough uses for the product that it can be shipped with additional features being added later. This is a great strategy for cash strapped companies to start turning a profit and paying back investors. However, as we look at CES 2018 and other events where every device and gadget is now connected to the internet, I have to question whether everything has been properly vetted and secured.

This topic is nothing new, security analysts have been discussing compromised IoT devices for ages, but it is an important topic. If companies are not focused on security from the start, it can be hard to bolt-on later. Security must come first, even with a minimum viable product.

Dan

Week 4 – Intel’s Vulnerability and It’s Impact

Earlier in the week, two severe vulnerabilities that affect Intel, AMD, and ARM processors were leaked to the media. These vulnerabilities – Spectre and Meltdown are serious and affect nearly every server, computer, and other device in the world. Intel seems to be taking the brunt of the publicity, likely because they own such a large percentage of the overall market.

The vulnerabilities were discovered by a team a Google and their write-up is here. Some key takeaways are that they successfully exploited the vulnerability on server processors, desktop processors, and even a mobile phone processor. Overall, this means that the vulnerability is far reaching and could be used in nearly any situation.

Russell Brandom of The Verge wrote an article earlier this week about how the largest impact will be felt by cloud providers. I could not agree more, but am also concerned about IoT devices and other devices that utilize x86 CPU. Many network infrastructure devices have x86 CPUs in them today. Firewalls, network switches, and routers all leverage chips that are now vulnerable and, in one case, can only be patched by replacing the chip.

This is problematic in many ways. Updating microcode on a processor requires network hardware to be rebooted, causing outage. Many organizations are going to have to take down important systems to patch the affected devices. Even still, that will only fix one of the vulnerabilities.

Originally, the CERT notification for Spectre and Meltdown listed CPU replacement as the solution. However, as noted by The Register, the notification has since been updated to include patches as solution.  The Register argues that chip manufacturers are in “denial” about their patches being a permanent fix and that chip replacement is still the most likely fix.

It will be interesting to watch this play out. While CSP’s rush to patch their respective clouds, I worry most about network infrastructure and IoT devices. Two platforms that rarely get updates, but will desperately need them soon.

Week 3 – Kali Linux and InSpy

Kali Linux is a useful tool for security practitioners and those that want to learn about cybersecurity. The Linux distribution, formerly known as BackTrack, was renamed Kalie Linux in 2013. The distribution contains hundreds of useful utilities and tools to help security professionals test their security posture.

In November 2017, the 2017.3 release of Kali Linux was uploaded for distribution. Four new tools were included in the update, one of which is InSpy.

InSpy is a useful social engineering tool. Focused on crawling LinkedIn, InSpy will gather intelligence about a companies job listings and employees. For security testers, this could be useful to ensure that HR departments are building job listings that meet security standards. It can also be helpful with determining employee compliance with social networking policies.

From a different perspective, InSpy could be really useful when doing reconnaissance on a business. Targeted phishing attempts could be launched by using the EmpSpy function of InSpy. Using this function, the tool will crawl LinkedIn for employees who work at a specific company and provided a list of employee names, titles, and email addresses.

With tools like this openly available, organizations must begin crafting a social media policy that includes special provisions for what information is shared. For instance, a policy could indicate that employees only use personal email addresses on LinkedIn and refrain from sharing which departments they work for.

InSpy and Kali Linux information can be found here – https://tools.kali.org/information-gathering/inspy

Week 2 – Finding Vulnerability Information Online

There are many sources available online to track down vulnerability and threat information. However, it is important to carefully consider the source and ensure that the information is credible. To aid in that, here are three websites known for providing valuable threat intelligence –

National Vulnerability Database

The National Vulnerability Database (NVD) is a database that contains information and analysis of CVEs. The database is maintained by the NIST Computer Security Division and was created in 2000.

In addition to the useful information provided within the database, the NVD website also houses several useful visualizations. Below is one example –

These visualizations are helpful in spotting trends that developers should be on the look out for. Authentication issues is currently on the rise, so it would make sense that development teams spend more time focusing on their authentication code.

Common Vulnerabilities and Exposures

Common Vulnerabilities and Exposures (CVE) have been around for many years. There have been over 98,000 CVE IDs created since the inception of the project. https://cve.mitre.org is the de facto home for the CVE database. This website provides information about each CVE ID. Users can search by CVE ID and the returned data will give you a description of the attack as well as several references for more information.

The CVE site also contains a page dedicated to other useful sites on the web. For instance, the NVD is mentioned as are other sites like US-CERT.

Talos Intelligence

Talos is the security organization that powers many of Cisco’s security products. The Talos group maintains an excellent web presence, providing visitors with in-depth blog posts as well as reports about vulnerabilities. These reports often cross-reference the CVE ID database.

In addition to their reports, the Talos group also maintains a blog. The Talos blog provides in-depth coverage of the latest attacks plaguing the internet. Because Talos is tasked with understanding threats and protecting customers, they are required to breakdown the attacks in order to protect against them. While doing that, they frequently write in-depth blog posts about the attack. Take this one about WannaCry as an example.

Week 10 – Aadhaar’s Security Problem

An interesting article popped up on Mashable.com this week about India’s biometric database Aadhaar. The Aadhaar database is India’s method to make up for the lack of birth certificates and other identification of Indian citizens. For many years, according to the article, the majority of people in India did not have birth certificates. To help combat this, the Indian Government designed Aadhaar.

Aadhaar is a biometric database which contains information on 99% of India. This database is used for more than just identification though. The government of India has plans to do away with credit cards, moving to fingerprint based transactions. In addition to being a payment gateway, Aadhaar is also poised to pivot into a digital wallet. India’s citizens will be able to load their health card and driver’s license into an electronic wallet of sorts, removing the need for normal cards.

This sounds great, but the problem is the database has never gone through an assessment or audit process. This, of course, has led to falsified entires into the database. These falsified entries are being used for all types of scams and a complete lack of oversight is leaving the citizens with little to no privacy.

My take – This is why we develop with a security first mindset. Aadhaar is already so big that it will be hard to transition into a more secure platform. If security would have been considered during software development, some of these problems could have been avoided.

Additionally, the government needs to consider routine audits of the database. A risk management strategy would really help the Indian government accelerate their chances of securing Aadhaar. Many identities are at risk if they do not adopt a risk management strategy for this database. A continuous assessment plan should be considered and adopted. This would certainly help with falsified records.

Week 9 – President Trump Fires White House CISO

In a follow-up to my Week 7 post about President Trump using a 5-year old Android phone, he has now decided to fire the CISO that President Obama hired in 2015. TechTarget Senior Reporter, Michael Heller, reports that the President is likely using his private security firm to handle duties until a replacement is hired.

When President Obama hired Cory Louie back in 2015, he wanted to help the White House better understand the risks and threats that they faced. While there was never much publicity about this hire, it is obvious that the former President and his staff felt the need to better protect the White House. I would tend to agree. As stated in my week 7 article, there are no laws dictating what types of technology the President can use, therefore it is easy to see why the President may want a security advisor. It is likely that Louie helped shape the President’s security strategy as well as making recommendations for how the President should use technology.

Frankly, if the President Trump’s private security team is allowing him to use a five year old phone, I have to question whether or not they can guide the White House away from security threats. I believe President Obama made the right decision in hiring a CISO for the White House, and am hopeful that President Trump will hire a replacement.

ICMP Redirects Broke the Voice Network – Understanding Traffic Flows

Some time ago Phil Gervasi tweeted about understanding traffic flows and it got me thinking about situations where I did not quite understand how the production network was functioning. Phil’s tweet was more about fault tolerance, but I think it applies to general Network Engineering. If you do not understand how data flows, it becomes very difficult to troubleshoot your network.

Continue reading →

2016 – Goals and Getting Things Done

It is a new year – 2016 is upon us! The blog got a facelift right before the new year and I renewed the SSL certificate a few days late (whoops…’Hey Siri, set a reminder to renew vlan50.com’s cert on 12/28/2016′). A facelift is not worth much if the site does not get updated more often (see Goal #1), so what better way to start than with a 2016 goals post?

Goals for 2016

Goal 1: Post More Often

Quite a generic goal isn’t it? The fact is a lot of things changed for me in 2015 and after starting off on the right track, I stopped posting around April. That was never my intention, time simply slipped away. I spent most of the summer developing four college courses, moved into a project management role in the fall, and then 2016 showed up.

Things have slowed down a bit, or perhaps I have gotten used to the new normal. Either way, in 2016 my goal is simply this – 26 posts by the time I write my 2017 Goals post (not counting this one). My expanded role has offered many writing opportunities and I have every intention of continuing my professional development as an Engineer. With that in mind, I should have ample chances to post quality content.

Goal 2: About That Professional Development

2015 was an interesting year in the Networking space. A lot of press would have you believe it was the year of SDN. I think you could argue that with Cisco’s ACI only having 1,000 customers, perhaps 2015 was a foundational year for SDN.

How does that relate to my professional development? SDN might not be here yet, but network automation is. As a result I plan to sharpen up my programming skills and put them to use, specifically focusing on Python.

There is more than that though. The way I got things done as an Engineer does not scale well with several other Engineers reporting to me. I will work to develop a project management skill set that translates and scales well. I believe in the Lean process and think that kanbans can be used effectively in Operations, just as in Development.

Finally I will be continue my journey down the Cisco Certification trail. Sometimes Engineers (myself included) put a little too much weight in the paper, but I believe Cisco Certifications are still a solid measuring stick for professional development.

So, how do I boil goal two into something measurable – a bulleted list!

  • Brush up those Python skills – build a tool to pull network inventory information
  • Project Management – Develop a style and measure its success
  • Cisco Certification – In the air, but I believe CCNP R&S would be a great goal for 2016.

Goal 3: Contribute More

In 2016 I plan on focusing some efforts in the open source community. Last year I wrote a series of scripts to help migrate from Observium to LibreNMS. That post has been wildly successful and I hope that I can continue to contribute to various open source tools that I use. The networking community is a special one, always willing to help each other and in 2016 I would like to focus a majority of my posts on helping other engineers.

Conclusion

2015 was an interesting year for me. With additional responsibilities and other projects taking a lot of time, I did not post nearly enough. Goal one is simply to rectify that – 26 posts by the time I write the 2017 goals post. Goal two and three will really facilitate the success of #1. Goals are great – they feel nice to set, even better to write about, but success is not defined by writing things down. Success is defined by doing things, 2016 will be the year of ‘getting things done.’