"Mesh Facade" by "Pedro Ângelo" on Flickr

Looking at the Nebula Overlay Meshed VPN Network from Slack

Around 2-3 years ago, Slack– the company who produces Slack the IM client, started working on a meshed overlay network product, called Nebula, to manage their environment. After two years of running their production network on the back of it, they decided to open source it. I found out about Nebula via a Medium Post that was mentioned in the HangOps Slack Group. I got interested in it, asked a few questions about Nebula in the Slack, and then in the Github Issues for it, and then recently raised a Pull Request to add more complete documentation than their single (heavily) commented config file.

So, let’s go into some details on why this is interesting to me.

1. Nebula uses a flat IPv4 network to identify all hosts in the network, no matter where in the network those hosts reside.

This means that I can address any host in my (self allocated) 198.18.0.0/16 network, and I don’t need to worry about routing tables, production/DR sites, network tromboneing and so on… it’s just… Flat.

[Note: Yes, that IP address “looks” like a public subnet – but it’s actually a testing network, allocated by IANA for network testing!]

2. Nebula has host-based firewalling built into the configuration file.

This means that once I know how my network should operate (yes, I know that’s a big ask), I can restrict my servers from being able to reach my laptops, or I can stop my web server from being able to talk to my database server, except for on the database ports. Lateral movement becomes a LOT harder.

This firewalling also looks a lot like (Network) Security Groups (for the AWS and Azure familiar), so you have a default “Deny” rule, and then layer “Allow” rules on top. You also have Inbound and Outbound rules, so if you want to stop your laptops from talking anything but DNS, SSH, HTTPS and ICMP over Nebula…. well, yep, you can do that :)

3. Nebula uses a PKI environment. Where you can have multiple Certificate Authorities.

This means that I have a central server running my certificate authority (CA), with a “backup” CA, stored offline – in case of dire disaster with my primary CA, and push both CA’s to all my nodes. If I suddenly need to replace all the certificates that my current CA signed, I can do that with minimal interruption to my nodes. Good stuff.

Nebula also created their own PKI attributes to identify the roles of each device in the Nebula environment. By signing that as part of the certificate on each node too, means your CA asserts that the role that certificate holds is valid for that node in the network.

Creating a node’s certificate is a simple command:

nebula-cert sign -name jon-laptop -ip 198.18.0.1/16 -groups admin,laptop,support

This certificate has the IP address of the node baked in (it’s 198.18.0.1) and the groups it’s part of (admin, laptop and support), as well as the host name (jon-laptop). I can use any of these three items in my firewall rules I mentioned above.

4. It creates a peer-to-peer, meshed VPN.

While it’s possible to create a peer-to-peer meshed VPN with commercial products, I’ve not seen any which are as light-weight to deploy as this. Each node finds all the other nodes in the network by using a collection of “Lighthouses” (similar to Torrent Seeds or Skype Super Nodes) which tells all the connecting nodes where all the other machines in the network are located. These then initiate UDP connections to the other nodes they want to talk to. If they are struggling (because of NAT or Double NAT), then there’s a NAT Punching process (called, humourously, “punchy”) which lets you signal via the Lighthouse that you’re trying to reach another node that can’t see your connection, and asks it to also connect out to you over UDP… thereby fixing the connection issue. All good.

5. Nebula has clients for Windows, Mac and Linux. Apparently there are clients for iOS in the works (meh, I’m not on Apple… but I know some are) and I’ve heard nothing about Android as yet, but as it’s on Linux, I’m sure some enterprising soul can take a look at it (the client is written in Go).

If you want to have a look at Nebula for your own testing, I’ve created a Terraform based environment on AWS and Azure to show how you’d manage it all using Ansible Tower, which builds:

2 VPCs (AWS) and 1 VNet (Azure)
6 subnets (3 public, 3 private)
1 public AWX (the upstream project from Ansible Tower) Server
1 private Nebula Certificate Authority
2 public Web Servers (one in AWS, one in Azure)
2 private Database Servers (one in AWS, one in Azure)
2 public Bastion Servers (one in AWS, one in Azure) – that lets AWX reach into the Private sections of the network, without exposing SSH from all the hosts.

If you don’t want to provision the Azure side, just remove load_web2_module.tf from the Terraform directory in that repo… Job’s a good’n!

I have plans to look at a couple of variables, like Nebula’s closest rival, ZeroTier, and to look at using SaltStack instead of Ansible, to reduce the need for the extra Bastion servers.

Featured image is “Mesh Facade” by “Pedro Ângelo” on Flickr and is released under a CC-BY-SA license.

“Swatch Water Store, Grand Central Station, NYC, 9/2016, pics by Mike Mozart of TheToyChannel and JeepersMedia on YouTube #Swatch #Watch” by “Mike Mozart” on Flickr

Time Based Security

I came across the concept of “Time Based Security” (TBS) in the Sysadministrivia podcast, S4E13.

I’m still digging into the details of it, but in essence, the “Armadillo” (Crunchy on the outside, soft on the inside) protection model is broken (sometimes known as the “Fortress Model”). You assume that your impenetrable network boundary will prevent attackers from getting to your sensitive data. While this may stop them for a while, what you’re actually seeing here is one part of a complex protection system, however many organisations miss the fact that this is just one part.

The examples used in the only online content I’ve found about this refer to a burglary.

In this context, your “Protection” (P) is measured in time. Perhaps you have hardened glass that takes 20 seconds to break.

Next, we evaluate “Detection” (D) which is also, surprisingly enough, measured in time. As the glass is hit, it triggers an alarm to a security facility. That takes 20 seconds to respond and goes to a dispatch centre, another 20 seconds for that to be answered and a police officer dispatched.

The police officer being dispatched is the “Response” (R). The police take (optimistically) 2 minutes to arrive (it was written in the 90’s so the police forces weren’t decimated then).

So, in the TBS system, we say that Detection (D) of 40 seconds plus Response (R) of 120 seconds = 160 seconds. This is greater than Protection (P) of 20 seconds, so we have an Exposure (E) time of 140 seconds E = P – (D + R). The question that is posed is, how much damage can be done in E?

So, compare this to your average pre-automation SOC. Your firewall, SIEM (Security Incident Event Management system), IDS (Intrusion Detection System) or WAF (Web Application Firewall) triggers an alarm. Someone is trying to do something (e.g. Denial Of Service attack, password spraying or port scanning for vulnerable services) a system you’re responsible for. While D might be in the tiny fractions of a minute (perhaps let’s say 1 minute, for maths sake), R is likely to be minutes or even hours, depending on the refresh rate of the ticket management system or alarm system (again, for maths sake, let’s say 60 minutes). So, D+R is now 61 minutes. How long is P really going to hold? Could it be less than 30 minutes against a determined attacker? (Let’s assume P is 30 minutes for maths sake).

Let’s do the calculation for a pre-automation SOC (Security Operations Centre). P-(D+R)=E. E here is 31 minutes. How much damage can an attacker do in 31 minutes? Could they put a backdoor into your system? Can they download sensitive data to a remote system? Could they pivot to your monitoring system, and remove the logs that said they were in there?

If you consider how much smaller the D and R numbers become with an event driven SOAR (Security Orchestration and Automation Response) system – does that improve your P and E numbers? Consider that if you can get E to 0, this could be considered to be “A Secure Environment”.

Also, consider the fact that many of the tools we implement for security reduce D and R, but if you’re not monitoring the outputs of the Detection components, then your response time grows significantly. If your Detection component is misconfigured in that it’s producing too many False Positives (for example, “The Boy Who Cried Wolf“), so you don’t see the real incident, then your Response might only be when a security service notifies you that your data, your service or your money has been exposed and lost. And that wouldn’t be good now… Time to look into automation 😁

Featured image is “Swatch Water Store, Grand Central Station, NYC, 9/2016, pics by Mike Mozart of TheToyChannel and JeepersMedia on YouTube #Swatch #Watch” by “Mike Mozart” on Flickr and is released under a CC-BY license.

JonTheNiceGuy and "The Chief" Peter Bleksley at BSides Liverpool 2019

Review of BSIDES Liverpool 2019

I had the privilege today to attend BSIDES Liverpool 2019. BSIDES is a infosec community conference. The majority of the talks were recorded, and I can strongly recommend making your way through the content when it becomes available.

Full disclosure: While my employer is a sponsor, I was not there to represent the company, I was just enjoying the show. A former colleague (good friend and, while he was still employed by Fujitsu, an FDE – so I think he still is one) is one of the organisers team.

The first talk I saw (aside from the welcome speech) was the keynote by Omri Segev Moyal (@gelossnake) about how to use serverless technologies (like AWS Lambda) to build a malware research platform. The key takeaway I have from that talk was how easy it is to build a simple python lambda script using Chalice. That was fantastic, and I’m looking forward to trying some things with that service!

For various reasons (mostly because I got talking to people), I missed the rest of the morning tracks except for the last talk before lunch. I heard great things about the Career Advice talk by Martin King, and the Social Engineering talk by Tom H, but will need to catch up on those on the videos released after.

Just before lunch we received a talk from “The Chief” (from the Channel 4 TV Series “Hunted”), Peter Bleksley, about an investigation he’s currently involved in. This was quite an intense session, and his history (the first 1/4 of his talk) was very interesting. Just before he went in for his talk, I got a selfie with him (which is the “Featured Image” for this post :) )

After lunch, I sat on the Rookies Track, and saw three fantastic talks, from Chrissi Robertson (@frootware) on Imposter Syndrome, Matt (@reversetor) on “Privacy in the age of Convenience” (reminding me of one of my very early talks at OggCamp/BarCamp Manchester) and Jan (@janfajfer) about detecting data leaks on mobile devices with EVPN. All three speakers were fab and nailed their content.

Next up was an unrecorded talk by Jamie (@2sec4u) about WannaCry, as he was part of the company who discovered the “Kill-Switch” domain. He gave a very detailed overview of the timeline about WannaCry, the current situation of the kill-switch, and a view on some of the data from infected-but-dormant machines which are still trying to reach the kill-switch. A very scary but well explained talk. Also, memes and rude words, but it’s clearly a subject that needed some levity, being part of a frankly rubbish set of circumstances.

After that was a talk from (two-out-of-six of) The Beer Farmers. This was a talk (mostly) about privacy and the lack of it from the social media systems of Facebook, Twitter and Google. As I listen to The Many Hats Club podcast, on which the Beer Farmers occasionally appear, it was a great experience matching faces to voices.

We finished the day on a talk by Finux (@f1nux) about Machiavelli as his writings (in the form of “The Prince”) would apply to Infosec. I was tempted to take a whole slew of photos of the slide deck, but figured I’d just wait for the video to be released, as it would, I’m sure, make more sense in context.

There was a closing talk, and then everyone retired to the bar. All in all, a great day, and I’m really glad I got the opportunity to go (thanks for your ticket Paul (@s7v7ns) – you missed out mate!)