Search results

AlgoSec ObjectFlow Optimizing policy automation through effective object management Download PDF Schedule time with one of our experts Schedule time with one of our experts Work email* First name* Last name* Company* country* Select country... Short answer* By submitting this form, I accept AlgoSec's privacy policy Continue

As you may have heard already, the National Security Agency (NSA) and the Federal Bureau of Investigation (FBI) released a joint... Cloud Security Drovorub’s Ability to Conceal C2 Traffic And Its Implications For Docker Containers Rony Moshkovich 2 min read Rony Moshkovich Short bio about author here Lorem ipsum dolor sit amet consectetur. Vitae donec tincidunt elementum quam laoreet duis sit enim. Duis mattis velit sit leo diam. Tags Share this article 8/15/20 Published As you may have heard already, the National Security Agency (NSA) and the Federal Bureau of Investigation (FBI) released a joint Cybersecurity Advisory about previously undisclosed Russian malware called Drovorub. According to the report, the malware is designed for Linux systems as part of its cyber espionage operations. Drovorub is a Linux malware toolset that consists of an implant coupled with a kernel module rootkit, a file transfer and port forwarding tool, and a Command and Control (C2) server. The name Drovorub originates from the Russian language. It is a complex word that consists of 2 roots (not the full words): “drov” and “rub” . The “o” in between is used to join both roots together. The root “drov” forms a noun “drova” , which translates to “firewood” , or “wood” . The root “rub” /ˈruːb/ forms a verb “rubit” , which translates to “to fell” , or “to chop” . Hence, the original meaning of this word is indeed a “woodcutter” . What the report omits, however, is that apart from the classic interpretation, there is also slang. In the Russian computer slang, the word “drova” is widely used to denote “drivers” . The word “rubit” also has other meanings in Russian. It may mean to kill, to disable, to switch off. In the Russian slang, “rubit” also means to understand something very well, to be professional in a specific field. It resonates with the English word “sharp” – to be able to cut through the problem. Hence, we have 3 possible interpretations of ‘ Drovorub ‘: someone who chops wood – “дроворуб” someone who disables other kernel-mode drivers – “тот, кто отрубает / рубит драйвера” someone who understands kernel-mode drivers very well – “тот, кто (хорошо) рубит в драйверах” Given that Drovorub does not disable other drivers, the last interpretation could be the intended one. In that case, “Drovorub” could be a code name of the project or even someone’s nickname. Let’s put aside the intricacies of the Russian translations and get a closer look into the report. DISCLAIMER Before we dive into some of the Drovorub analysis aspects, we need to make clear that neither FBI nor NSA has shared any hashes or any samples of Drovorub. Without the samples, it’s impossible to conduct a full reverse engineering analysis of the malware. Netfilter Hiding According to the report, the Drovorub-kernel module registers a Netfilter hook. A network packet filter with a Netfilter hook ( NF_INET_LOCAL_IN and NF_INET_LOCAL_OUT ) is a common malware technique. It allows a backdoor to watch passively for certain magic packets or series of packets, to extract C2 traffic. What is interesting though, is that the driver also hooks the kernel’s nf_register_hook() function. The hook handler will register the original Netfilter hook, then un-register it, then re-register the kernel’s own Netfilter hook. According to the nf_register_hook() function in the Netfilter’s source , if two hooks have the same protocol family (e.g., PF_INET ), and the same hook identifier (e.g., NF_IP_INPUT ), the hook execution sequence is determined by priority. The hook list enumerator breaks at the position of an existing hook with a priority number elem->priority higher than the new hook’s priority number reg->priority : int nf_register_hook ( struct nf_hook_ops * reg) { struct nf_hook_ops * elem; int err; err = mutex_lock_interruptible( & nf_hook_mutex); if (err < 0 ) return err; list_for_each_entry(elem, & nf_hooks[reg -> pf][reg -> hooknum], list) { if (reg -> priority < elem -> priority) break ; } list_add_rcu( & reg -> list, elem -> list.prev); mutex_unlock( & nf_hook_mutex); ... return 0 ; } In that case, the new hook is inserted into the list, so that the higher-priority hook’s PREVIOUS link would point into the newly inserted hook. What happens if the new hook’s priority is also the same, such as NF_IP_PRI_FIRST – the maximum hook priority? In that case, the break condition will not be met, the list iterator list_for_each_entry will slide past the existing hook, and the new hook will be inserted after it as if the new hook’s priority was higher. By re-inserting its Netfilter hook in the hook handler of the nf_register_hook() function, the driver makes sure the Drovorub’s Netfilter hook will beat any other registered hook at the same hook number and with the same (maximum) priority. If the intercepted TCP packet does not belong to the hidden TCP connection, or if it’s destined to or originates from another process, hidden by Drovorub’s kernel-mode driver, the hook will return 5 ( NF_STOP ). Doing so will prevent other hooks from being called to process the same packet. Security Implications For Docker Containers Given that Drovorub toolset targets Linux and contains a port forwarding tool to route network traffic to other hosts on the compromised network, it would not be entirely unreasonable to assume that this toolset was detected in a client’s cloud infrastructure. According to Gartner’s prediction , in just two years, more than 75% of global organizations will be running cloud-native containerized applications in production, up from less than 30% today. Would the Drovorub toolset survive, if the client’s cloud infrastructure was running containerized applications? Would that facilitate the attack or would it disrupt it? Would it make the breach stealthier? To answer these questions, we have tested a different malicious toolset, CloudSnooper, reported earlier this year by Sophos. Just like Drovorub, CloudSnooper’s kernel-mode driver also relies on a Netfilter hook ( NF_INET_LOCAL_IN and NF_INET_LOCAL_OUT ) to extract C2 traffic from the intercepted TCP packets. As seen in the FBI/NSA report, the Volatility framework was used to carve the Drovorub kernel module out of the host, running CentOS. In our little lab experiment, let’s also use CentOS host. To build a new Docker container image, let’s construct the following Dockerfile: FROM scratch ADD centos-7.4.1708-docker.tar.xz / ADD rootkit.ko / CMD [“/bin/bash”] The new image, built from scratch, will have the CentOS 7.4 installed. The kernel-mode rootkit will be added to its root directory. Let’s build an image from our Dockerfile, and call it ‘test’: [root@localhost 1]# docker build . -t test Sending build context to Docker daemon 43.6MB Step 1/4 : FROM scratch —> Step 2/4 : ADD centos-7.4.1708-docker.tar.xz / —> 0c3c322f2e28 Step 3/4 : ADD rootkit.ko / —> 5aaa26212769 Step 4/4 : CMD [“/bin/bash”] —> Running in 8e34940342a2 Removing intermediate container 8e34940342a2 —> 575e3875cdab Successfully built 575e3875cdab Successfully tagged test:latest Next, let’s execute our image interactively (with pseudo-TTY and STDIN ): docker run -it test The executed image will be waiting for our commands: [root@8921e4c7d45e /]# Next, let’s try to load the malicious kernel module: [root@8921e4c7d45e /]# insmod rootkit.ko The output of this command is: insmod: ERROR: could not insert module rootkit.ko: Operation not permitted The reason why it failed is that by default, Docker containers are ‘unprivileged’. Loading a kernel module from a docker container requires a special privilege that allows it doing so. Let’s repeat our experiment. This time, let’s execute our image either in a fully privileged mode or by enabling only one capability – a capability to load and unload kernel modules ( SYS_MODULE ). docker run -it –privileged test or docker run -it –cap-add SYS_MODULE test Let’s load our driver again: [root@547451b8bf87 /]# insmod rootkit.ko This time, the command is executed silently. Running lsmod command allows us to enlist the driver and to prove it was loaded just fine. A little magic here is to quit the docker container and then delete its image: docker rmi -f test Next, let’s execute lsmod again, only this time on the host. The output produced by lsmod will confirm the rootkit module is loaded on the host even after the container image is fully unloaded from memory and deleted! Let’s see what ports are open on the host: [root@localhost 1]# netstat -tulpn Active Internet connections (only servers) Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name tcp 0 0 0.0.0.0:22 0.0.0.0:* LISTEN 1044/sshd With the SSH server running on port 22 , let’s send a C2 ‘ping’ command to the rootkit over port 22 : [root@localhost 1]# python client.py 127.0.0.1 22 8080 rrootkit-negotiation: hello The ‘hello’ response from the rootkit proves it’s fully operational. The Netfilter hook detects a command concealed in a TCP packet transferred over port 22 , even though the host runs SSH server on port 22 . How was it possible that a rootkit loaded from a docker container ended up loaded on the host? The answer is simple: a docker container is not a virtual machine. Despite the namespace and ‘control groups’ isolation, it still relies on the same kernel as the host. Therefore, a kernel-mode rootkit loaded from inside a Docker container instantly compromises the host, thus allowing the attackers to compromise other containers that reside on the same host. It is true that by default, a Docker container is ‘unprivileged’ and hence, may not load kernel-mode drivers. However, if a host is compromised, or if a trojanized container image detects the presence of the SYS_MODULE capability (as required by many legitimate Docker containers), loading a kernel-mode rootkit on a host from inside a container becomes a trivial task. Detecting the SYS_MODULE capability ( cap_sys_module ) from inside the container: [root@80402f9c2e4c /]# capsh –print Current: = cap_chown, … cap_sys_module, … Conclusion This post is drawing a parallel between the recently reported Drovorub rootkit and CloudSnooper, a rootkit reported earlier this year. Allegedly built by different teams, both of these Linux rootkits have one mechanism in common: a Netfilter hook ( NF_INET_LOCAL_IN and NF_INET_LOCAL_OUT ) and a toolset that enables tunneling of the traffic to other hosts within the same compromised cloud infrastructure. We are still hunting for the hashes and samples of Drovorub. Unfortunately, the YARA rules published by FBI/NSA cause False Positives. For example, the “Rule to detect Drovorub-server, Drovorub-agent, and Drovorub-client binaries based on unique strings and strings indicating statically linked libraries” enlists the following strings: “Poco” “Json” “OpenSSL” “clientid” “—–BEGIN” “—–END” “tunnel” The string “Poco” comes from the POCO C++ Libraries that are used for over 15 years. It is w-a-a-a-a-y too generic, even in combination with other generic strings. As a result, all these strings, along with the ELF header and a file size between 1MB and 10MB, produce a false hit on legitimate ARM libraries, such as a library used for GPS navigation on Android devices: f058ebb581f22882290b27725df94bb302b89504 56c36bfd4bbb1e3084e8e87657f02dbc4ba87755 Nevertheless, based on the information available today, our interest is naturally drawn to the security implications of these Linux rootkits for the Docker containers. Regardless of what security mechanisms may have been compromised, Docker containers contribute an additional attack surface, another opportunity for the attackers to compromise the hosts and other containers within the same organization. The scenario outlined in this post is purely hypothetical. There is no evidence that supports that Drovorub may have affected any containers. However, an increase in volume and sophistication of attacks against Linux-based cloud-native production environments, coupled with the increased proliferation of containers, suggests that such a scenario may, in fact, be plausible. Schedule a demo Related Articles 2025 in review: What innovations and milestones defined AlgoSec’s transformative year in 2025? AlgoSec Reviews Mar 19, 2023 · 2 min read Navigating Compliance in the Cloud AlgoSec Cloud Mar 19, 2023 · 2 min read 5 Multi-Cloud Environments Cloud Security Mar 19, 2023 · 2 min read Speak to one of our experts Speak to one of our experts Work email* First name* Last name* Company* country* Select country... Short answer* By submitting this form, I accept AlgoSec's privacy policy Schedule a call

Network management & policy change automation Download PDF Schedule time with one of our experts Schedule time with one of our experts Work email* First name* Last name* Company* country* Select country... Short answer* By submitting this form, I accept AlgoSec's privacy policy Continue

HKMA Compliance Download PDF Schedule time with one of our experts Schedule time with one of our experts Work email* First name* Last name* Company* country* Select country... Short answer* By submitting this form, I accept AlgoSec's privacy policy Continue

AlgoSec for GDPR Download PDF Schedule time with one of our experts Schedule time with one of our experts Work email* First name* Last name* Company* country* Select country... Short answer* By submitting this form, I accept AlgoSec's privacy policy Continue

Business-driven Security Management For Local Governments Download PDF Schedule time with one of our experts Schedule time with one of our experts Work email* First name* Last name* Company* country* Select country... Short answer* By submitting this form, I accept AlgoSec's privacy policy Continue

Like all modern cloud providers, Amazon adopts the shared responsibility model for cloud security. Amazon guarantees secure... AWS NACL best practices: How to combine security groups with network ACLs effectively Prof. Avishai Wool 2 min read Prof. Avishai Wool Short bio about author here Lorem ipsum dolor sit amet consectetur. Vitae donec tincidunt elementum quam laoreet duis sit enim. Duis mattis velit sit leo diam. Tags Share this article 8/28/23 Published Like all modern cloud providers, Amazon adopts the shared responsibility model for cloud security. Amazon guarantees secure infrastructure for Amazon Web Services, while AWS users are responsible for maintaining secure configurations. That requires using multiple AWS services and tools to manage traffic. You’ll need to develop a set of inbound rules for incoming connections between your Amazon Virtual Private Cloud (VPC) and all of its Elastic Compute (EC2) instances and the rest of the Internet. You’ll also need to manage outbound traffic with a series of outbound rules. Your Amazon VPC provides you with several tools to do this. The two most important ones are security groups and Network Access Control Lists (NACLs). Security groups are stateful firewalls that secure inbound traffic for individual EC2 instances. Network ACLs are stateless firewalls that secure inbound and outbound traffic for VPC subnets. Managing AWS VPC security requires configuring both of these tools appropriately for your unique security risk profile. This means planning your security architecture carefully to align it the rest of your security framework. For example, your firewall rules impact the way Amazon Identity Access Management (IAM) handles user permissions. Some (but not all) IAM features can be implemented at the network firewall layer of security. Before you can manage AWS network security effectively , you must familiarize yourself with how AWS security tools work and what sets them apart. Everything you need to know about security groups vs NACLs AWS security groups explained: Every AWS account has a single default security group assigned to the default VPC in every Region. It is configured to allow inbound traffic from network interfaces assigned to the same group, using any protocol and any port. It also allows all outbound traffic using any protocol and any port. Your default security group will also allow all outbound IPv6 traffic once your VPC is associated with an IPv6 CIDR block. You can’t delete the default security group, but you can create new security groups and assign them to AWS EC2 instances. Each security group can only contain up to 60 rules, but you can set up to 2500 security groups per Region. You can associate many different security groups to a single instance, potentially combining hundreds of rules. These are all allow rules that allow traffic to flow according the ports and protocols specified. For example, you might set up a rule that authorizes inbound traffic over IPv6 for linux SSH commands and sends it to a specific destination. This could be different from the destination you set for other TCP traffic. Security groups are stateful, which means that requests sent from your instance will be allowed to flow regardless of inbound traffic rules. Similarly, VPC security groups automatically responses to inbound traffic to flow out regardless of outbound rules. However, since security groups do not support deny rules, you can’t use them to block a specific IP address from connecting with your EC2 instance. Be aware that Amazon EC2 automatically blocks email traffic on port 25 by default – but this is not included as a specific rule in your default security group. AWS NACLs explained: Your VPC comes with a default NACL configured to automatically allow all inbound and outbound network traffic. Unlike security groups, NACLs filter traffic at the subnet level. That means that Network ACL rules apply to every EC2 instance in the subnet, allowing users to manage AWS resources more efficiently. Every subnet in your VPC must be associated with a Network ACL. Any single Network ACL can be associated with multiple subnets, but each subnet can only be assigned to one Network ACL at a time. Every rule has its own rule number, and Amazon evaluates rules in ascending order. The most important characteristic of NACL rules is that they can deny traffic. Amazon evaluates these rules when traffic enters or leaves the subnet – not while it moves within the subnet. You can access more granular data on data flows using VPC flow logs. Since Amazon evaluates NACL rules in ascending order, make sure that you place deny rules earlier in the table than rules that allow traffic to multiple ports. You will also have to create specific rules for IPv4 and IPv6 traffic – AWS treats these as two distinct types of traffic, so rules that apply to one do not automatically apply to the other. Once you start customizing NACLs, you will have to take into account the way they interact with other AWS services. For example, Elastic Load Balancing won’t work if your NACL contains a deny rule excluding traffic from 0.0.0.0/0 or the subnet’s CIDR. You should create specific inclusions for services like Elastic Load Balancing, AWS Lambda, and AWS CloudWatch. You may need to set up specific inclusions for third-party APIs, as well. You can create these inclusions by specifying ephemeral port ranges that correspond to the services you want to allow. For example, NAT gateways use ports 1024 to 65535. This is the same range covered by AWS Lambda functions, but it’s different than the range used by Windows operating systems. When creating these rules, remember that unlike security groups, NACLs are stateless. That means that when responses to allowed traffic are generated, those responses are subject to NACL rules. Misconfigured NACLs deny traffic responses that should be allowed, leading to errors, reduced visibility, and potential security vulnerabilities . How to configure and map NACL associations A major part of optimizing NACL architecture involves mapping the associations between security groups and NACLs. Ideally, you want to enforce a specific set of rules at the subnet level using NACLs, and a different set of instance-specific rules at the security group level. Keeping these rulesets separate will prevent you from setting inconsistent rules and accidentally causing unpredictable performance problems. The first step in mapping NACL associations is using the Amazon VPC console to find out which NACL is associated with a particular subnet. Since NACLs can be associated with multiple subnets, you will want to create a comprehensive list of every association and the rules they contain. To find out which NACL is associated with a subnet: Open the Amazon VPC console . Select Subnets in the navigation pane. Select the subnet you want to inspect. The Network ACL tab will display the ID of the ACL associated with that network, and the rules it contains. To find out which subnets are associated with a NACL: Open the Amazon VPC console . Select Network ACLS in the navigation pane. Click over to the column entitled Associated With. Select a Network ACL from the list. Look for Subnet associations on the details pane and click on it. The pane will show you all subnets associated with the selected Network ACL. Now that you know how the difference between security groups and NACLs and you can map the associations between your subnets and NACLs, you’re ready to implement some security best practices that will help you strengthen and simplify your network architecture. 5 best practices for AWS NACL management Pay close attention to default NACLs, especially at the beginning Since every VPC comes with a default NACL, many AWS users jump straight into configuring their VPC and creating subnets, leaving NACL configuration for later. The problem here is that every subnet associated with your VPC will inherit the default NACL. This allows all traffic to flow into and out of the network. Going back and building a working security policy framework will be difficult and complicated – especially if adjustments are still being made to your subnet-level architecture. Taking time to create custom NACLs and assign them to the appropriate subnets as you go will make it much easier to keep track of changes to your security posture as you modify your VPC moving forward. Implement a two-tiered system where NACLs and security groups complement one another Security groups and NACLs are designed to complement one another, yet not every AWS VPC user configures their security policies accordingly. Mapping out your assets can help you identify exactly what kind of rules need to be put in place, and may help you determine which tool is the best one for each particular case. For example, imagine you have a two-tiered web application with web servers in one security group and a database in another. You could establish inbound NACL rules that allow external connections to your web servers from anywhere in the world (enabling port 443 connections) while strictly limiting access to your database (by only allowing port 3306 connections for MySQL). Look out for ineffective, redundant, and misconfigured deny rules Amazon recommends placing deny rules first in the sequential list of rules that your NACL enforces. Since you’re likely to enforce multiple deny rules per NACL (and multiple NACLs throughout your VPC), you’ll want to pay close attention to the order of those rules, looking for conflicts and misconfigurations that will impact your security posture. Similarly, you should pay close attention to the way security group rules interact with your NACLs. Even misconfigurations that are harmless from a security perspective may end up impacting the performance of your instance, or causing other problems. Regularly reviewing your rules is a good way to prevent these mistakes from occurring. Limit outbound traffic to the required ports or port ranges When creating a new NACL, you have the ability to apply inbound or outbound restrictions. There may be cases where you want to set outbound rules that allow traffic from all ports. Be careful, though. This may introduce vulnerabilities into your security posture. It’s better to limit access to the required ports, or to specify the corresponding port range for outbound rules. This establishes the principle of least privilege to outbound traffic and limits the risk of unauthorized access that may occur at the subnet level. Test your security posture frequently and verify the results How do you know if your particular combination of security groups and NACLs is optimal? Testing your architecture is a vital step towards making sure you haven’t left out any glaring vulnerabilities. It also gives you a good opportunity to address misconfiguration risks. This doesn’t always mean actively running penetration tests with experienced red team consultants, although that’s a valuable way to ensure best-in-class security. It also means taking time to validate your rules by running small tests with an external device. Consider using AWS flow logs to trace the way your rules direct traffic and using that data to improve your work. How to diagnose security group rules and NACL rules with flow logs Flow logs allow you to verify whether your firewall rules follow security best practices effectively. You can follow data ingress and egress and observe how data interacts with your AWS security rule architecture at each step along the way. This gives you clear visibility into how efficient your route tables are, and may help you configure your internet gateways for optimal performance. Before you can use the Flow Log CLI, you will need to create an IAM role that includes a policy granting users the permission to create, configure, and delete flow logs. Flow logs are available at three distinct levels, each accessible through its own console: Network interfaces VPCs Subnets You can use the ping command from an external device to test the way your instance’s security group and NACLs interact. Your security group rules (which are stateful) will allow the response ping from your instance to go through. Your NACL rules (which are stateless) will not allow the outbound ping response to travel back to your device. You can look for this activity through a flow log query. Here is a quick tutorial on how to create a flow log query to check your AWS security policies. First you’ll need to create a flow log in the AWS CLI. This is an example of a flow log query that captures all rejected traffic for a specified network interface. It delivers the flow logs to a CloudWatch log group with permissions specified in the IAM role: aws ec2 create-flow-logs \ –resource-type NetworkInterface \ –resource-ids eni-1235b8ca123456789 \ –traffic-type ALL \ –log-group-name my-flow-logs \ –deliver-logs-permission-arn arn:aws:iam::123456789101:role/publishFlowLogs Assuming your test pings represent the only traffic flowing between your external device and EC2 instance, you’ll get two records that look like this: 2 123456789010 eni-1235b8ca123456789 203.0.113.12 172.31.16.139 0 0 1 4 336 1432917027 1432917142 ACCEPT OK 2 123456789010 eni-1235b8ca123456789 172.31.16.139 203.0.113.12 0 0 1 4 336 1432917094 1432917142 REJECT OK To parse this data, you’ll need to familiarize yourself with flow log syntax. Default flow log records contain 14 arguments, although you can also expand custom queries to return more than double that number: Version tells you the version currently in use. Default flow logs requests use Version 2. Expanded custom requests may use Version 3 or 4. Account-id tells you the account ID of the owner of the network interface that traffic is traveling through. The record may display as unknown if the network interface is part of an AWS service like a Network Load Balancer. Interface-id shows the unique ID of the network interface for the traffic currently under inspection. Srcaddr shows the source of incoming traffic, or the address of the network interface for outgoing traffic. In the case of IPv4 addresses for network interfaces, it is always its private IPv4 address. Dstaddr shows the destination of outgoing traffic, or the address of the network interface for incoming traffic. In the case of IPv4 addresses for network interfaces, it is always its private IPv4 address. Srcport is the source port for the traffic under inspection. Dstport is the destination port for the traffic under inspection. Protocol refers to the corresponding IANA traffic protocol number . Packets describes the number of packets transferred. Bytes describes the number of bytes transferred. Start shows the start time when the first data packet was received. This could be up to one minute after the network interface transmitted or received the packet. End shows the time when the last data packet was received. This can be up to one minutes after the network interface transmitted or received the data packet. Action describes what happened to the traffic under inspection: ACCEPT means that traffic was allowed to pass. REJECT means the traffic was blocked, typically by security groups or NACLs. Log-status confirms the status of the flow log: OK means data is logging normally. NODATA means no network traffic to or from the network interface was detected during the specified interval. SKIPDATA means some flow log records are missing, usually due to internal capacity restraints or other errors. Going back to the example above, the flow log output shows that a user sent a command from a device with the IP address 203.0.113.12 to the network interface’s private IP address, which is 172.31.16.139. The security group’s inbound rules allowed the ICMP traffic to travel through, producing an ACCEPT record. However, the NACL did not let the ping response go through, because it is stateless. This generated the REJECT record that followed immediately after. If you configure your NACL to permit output ICMP traffic and run this test again, the second flow log record will change to ACCEPT. azon Web Services (AWS) is one of the most popular options for organizations looking to migrate their business applications to the cloud. It’s easy to see why: AWS offers high capacity, scalable and cost-effective storage, and a flexible, shared responsibility approach to security. Essentially, AWS secures the infrastructure, and you secure whatever you run on that infrastructure. However, this model does throw up some challenges. What exactly do you have control over? How can you customize your AWS infrastructure so that it isn’t just secure today, but will continue delivering robust, easily managed security in the future? The basics: security groups AWS offers virtual firewalls to organizations, for filtering traffic that crosses their cloud network segments. The AWS firewalls are managed using a concept called Security Groups. These are the policies, or lists of security rules, applied to an instance – a virtualized computer in the AWS estate. AWS Security Groups are not identical to traditional firewalls, and they have some unique characteristics and functionality that you should be aware of, and we’ve discussed them in detail in video lesson 1: the fundamentals of AWS Security Groups , but the crucial points to be aware of are as follows. First, security groups do not deny traffic – that is, all the rules in security groups are positive, and allow traffic. Second, while security group rules can be set to specify a traffic source, or a destination, they cannot specify both on the same rule. This is because AWS always sets the unspecified side (source or destination) as the instance to which the group is applied. Finally, single security groups can be applied to multiple instances, or multiple security groups can be applied to a single instance: AWS is very flexible. This flexibility is one of the unique benefits of AWS, allowing organizations to build bespoke security policies across different functions and even operating systems, mixing and matching them to suit their needs. Adding Network ACLs into the mix To further enhance and enrich its security filtering capabilities AWS also offers a feature called Network Access Control Lists (NACLs). Like security groups, each NACL is a list of rules, but there are two important differences between NACLs and security groups. The first difference is that NACLs are not directly tied to instances, but are tied with the subnet within your AWS virtual private cloud that contains the relevant instance. This means that the rules in a NACL apply to all of the instances within the subnet, in addition to all the rules from the security groups. So a specific instance inherits all the rules from the security groups associated with it, plus the rules associated with a NACL which is optionally associated with a subnet containing that instance. As a result NACLs have a broader reach, and affect more instances than a security group does. The second difference is that NACLs can be written to include an explicit action, so you can write ‘deny’ rules – for example to block traffic from a particular set of IP addresses which are known to be compromised. The ability to write ‘deny’ actions is a crucial part of NACL functionality. It’s all about the order As a consequence, when you have the ability to write both ‘allow’ rules and ‘deny’ rules, the order of the rules now becomes important. If you switch the order of the rules between a ‘deny’ and ‘allow’ rule, then you’re potentially changing your filtering policy quite dramatically. To manage this, AWS uses the concept of a ‘rule number’ within each NACL. By specifying the rule number, you can identify the correct order of the rules for your needs. You can choose which traffic you deny at the outset, and which you then actively allow. As such, with NACLs you can manage security tasks in a way that you cannot do with security groups alone. However, we did point out earlier that an instance inherits security rules from both the security groups, and from the NACLs – so how do these interact? The order by which rules are evaluated is this; For inbound traffic, AWS’s infrastructure first assesses the NACL rules. If traffic gets through the NACL, then all the security groups that are associated with that specific instance are evaluated, and the order in which this happens within and among the security groups is unimportant because they are all ‘allow’ rules. For outbound traffic, this order is reversed: the traffic is first evaluated against the security groups, and then finally against the NACL that is associated with the relevant subnet. You can see me explain this topic in person in my new whiteboard video: Schedule a demo Related Articles 2025 in review: What innovations and milestones defined AlgoSec’s transformative year in 2025? AlgoSec Reviews Mar 19, 2023 · 2 min read Navigating Compliance in the Cloud AlgoSec Cloud Mar 19, 2023 · 2 min read 5 Multi-Cloud Environments Cloud Security Mar 19, 2023 · 2 min read Speak to one of our experts Speak to one of our experts Work email* First name* Last name* Company* country* Select country... Short answer* By submitting this form, I accept AlgoSec's privacy policy Schedule a call

Previous Part of the analysis is available here. Next Part of the analysis is available here. Update from 19 December 2020: ‍Prevasio... Cloud Security Sunburst Backdoor, Part II: DGA & The List of Victims Rony Moshkovich 2 min read Rony Moshkovich Short bio about author here Lorem ipsum dolor sit amet consectetur. Vitae donec tincidunt elementum quam laoreet duis sit enim. Duis mattis velit sit leo diam. Tags Share this article 12/17/20 Published Previous Part of the analysis is available here . Next Part of the analysis is available here . Update from 19 December 2020: Prevasio would like to thank Zetalytics for providing us with an updated (larger) list of passive (historic) DNS queries for the domains generated by the malware. As described in the first part of our analysis, the DGA (Domain Generation Algorithm) of the Sunburst backdoor produces a domain name that may look like: fivu4vjamve5vfrtn2huov[.]appsync-api.us-west-2[.]avsvmcloud[.]com The first part of the domain name (before the first dot) consists of a 16-character random string, appended with an encoded computer’s domain name. This is the domain in which the local computer is registered. From the example string above, we can conclude that the encoded computer’s domain starts from the 17th character and up until the dot (highlighted in yellow): fivu4vjamve5vfrt n2huov In order to encode a local computer’s domain name, the malware uses one of 2 simple methods: Method 1 : a substitution table, if the domain name consists of small letters, digits, or special characters ‘-‘, ‘_’, ‘.’ Method 2 : base64 with a custom alphabet, in case of capital letters present in the domain name Method 1 In our example, the encoded domain name is “n2huov” . As it does not have any capital letters, the malware encodes it with a substitution table “rq3gsalt6u1iyfzop572d49bnx8cvmkewhj” . For each character in the domain name, the encoder replaces it with a character located in the substitution table four characters right from the original character. In order to decode the name back, all we have to do is to replace each encoded character with another character, located in the substitution table four characters left from the original character. To illustrate this method, imagine that the original substitution table is printed on a paper strip and then covered with a card with 6 perforated windows. Above each window, there is a sticker note with a number on it, to reflect the order of characters in the word “n2huov” , where ‘n’ is #1, ‘2’ is #2, ‘h’ is #3 and so on: Once the paper strip is pulled by 4 characters right, the perforated windows will reveal a different word underneath the card: “domain” , where ‘d’ is #1, ‘o’ is #2, ‘m’ is #3, etc.: A special case is reserved for such characters as ‘0’ , ‘-‘ , ‘_’ , ‘.’ . These characters are encoded with ‘0’ , followed with a character from the substitution table. An index of that character in the substitution table, divided by 4, provides an index within the string “0_-.” . The following snippet in C# illustrates how an encoded string can be decoded: static string decode_domain( string s) { string table = "rq3gsalt6u1iyfzop572d49bnx8cvmkewhj" ; string result = "" ; for ( int i = 0 ; i < s.Length; i++) { if (s[i] != '0' ) { result += table[(table.IndexOf(s[i]) + table.Length - 4 ) % table.Length]; } else { if (i < s.Length - 1 ) { if (table.Contains(s[i + 1 ])) { result += "0_-." [table.IndexOf(s[i + 1 ]) % 4 ]; } else { break ; } } i++; } } return result; } Method 2 This method is a standard base64 encoder with a custom alphabet “ph2eifo3n5utg1j8d94qrvbmk0sal76c” . Here is a snippet in C# that provides a decoder: public static string FromBase32String( string str) { string table = "ph2eifo3n5utg1j8d94qrvbmk0sal76c" ; int numBytes = str.Length * 5 / 8 ; byte [] bytes = new Byte[numBytes]; int bit_buffer; int currentCharIndex; int bits_in_buffer; if (str.Length < 3 ) { bytes[ 0 ] = ( byte )(table.IndexOf(str[ 0 ]) | table.IndexOf(str[ 1 ]) << 5 ); return System.Text.Encoding.UTF8.GetString(bytes); } bit_buffer = (table.IndexOf(str[ 0 ]) | table.IndexOf(str[ 1 ]) << 5 ); bits_in_buffer = 10 ; currentCharIndex = 2 ; for ( int i = 0 ; i < bytes.Length; i++) { bytes[i] = ( byte )bit_buffer; bit_buffer >>= 8 ; bits_in_buffer -= 8 ; while (bits_in_buffer < 8 && currentCharIndex < str.Length) { bit_buffer |= table.IndexOf(str[currentCharIndex++]) << bits_in_buffer; bits_in_buffer += 5 ; } } return System.Text.Encoding.UTF8.GetString(bytes); } When the malware encodes a domain using Method 2, it prepends the encrypted string with a double zero character: “00” . Following that, extracting a domain part of an encoded domain name (long form) is as simple as: static string get_domain_part( string s) { int i = s.IndexOf( ".appsync-api" ); if (i > 0 ) { s = s.Substring( 0 , i); if (s.Length > 16 ) { return s.Substring( 16 ); } } return "" ; } Once the domain part is extracted, the decoded domain name can be obtained by using Method 1 or Method 2, as explained above: if (domain.StartsWith( "00" )) { decoded = FromBase32String(domain.Substring( 2 )); } else { decoded = decode_domain(domain); } Decrypting the Victims’ Domain Names To see the decoder in action, let’s select 2 lists: List #1 Bambenek Consulting has provided a list of observed hostnames for the DGA domain. List #2 The second list has surfaced in a Paste bin paste , allegedly sourced from Zetalytics / Zonecruncher . NOTE: This list is fairly ‘noisy’, as it has non-decodable domain names. By feeding both lists to our decoder, we can now obtain a list of decoded domains, that could have been generated by the victims of the Sunburst backdoor. DISCLAIMER: It is not clear if the provided lists contain valid domain names that indeed belong to the victims. It is quite possible that the encoded domain names were produced by third-party tools, sandboxes, or by researchers that investigated and analysed the backdoor. The decoded domain names are provided purely as a reverse engineering exercise. The resulting list was manually processed to eliminate noise, and to exclude duplicate entries. Following that, we have made an attempt to map the obtained domain names to the company names, using Google search. Reader’s discretion is advised as such mappings could be inaccurate. Decoded Domain Mapping (Could Be Inaccurate) hgvc.com Hilton Grand Vacations Amerisaf AMERISAFE, Inc. kcpl.com Kansas City Power and Light Company SFBALLET San Francisco Ballet scif.com State Compensation Insurance Fund LOGOSTEC Logostec Ventilação Industrial ARYZTA.C ARYZTA Food Solutions bmrn.com BioMarin Pharmaceutical Inc. AHCCCS.S Arizona Health Care Cost Containment System nnge.org Next Generation Global Education cree.com Cree, Inc (semiconductor products) calsb.org The State Bar of California rbe.sk.ca Regina Public Schools cisco.com Cisco Systems pcsco.com Professional Computer Systems barrie.ca City of Barrie ripta.com Rhode Island Public Transit Authority uncity.dk UN City (Building in Denmark) bisco.int Boambee Industrial Supplies (Bisco) haifa.edu University of Haifa smsnet.pl SMSNET, Poland fcmat.org Fiscal Crisis and Management Assistance Team wiley.com Wiley (publishing) ciena.com Ciena (networking systems) belkin.com Belkin spsd.sk.ca Saskatoon Public Schools pqcorp.com PQ Corporation ftfcu.corp First Tech Federal Credit Union bop.com.pk The Bank of Punjab nvidia.com NVidia insead.org INSEAD (non-profit, private university) usd373.org Newton Public Schools agloan.ads American AgCredit pageaz.gov City of Page jarvis.lab Erich Jarvis Lab ch2news.tv Channel 2 (Israeli TV channel) bgeltd.com Bradford / Hammacher Remote Support Software dsh.ca.gov California Department of State Hospitals dotcomm.org Douglas Omaha Technology Commission sc.pima.gov Arizona Superior Court in Pima County itps.uk.net IT Professional Services, UK moncton.loc City of Moncton acmedctr.ad Alameda Health System csci-va.com Computer Systems Center Incorporated keyano.local Keyano College uis.kent.edu Kent State University alm.brand.dk Sydbank Group (Banking, Denmark) ironform.com Ironform (metal fabrication) corp.ncr.com NCR Corporation ap.serco.com Serco Asia Pacific int.sap.corp SAP mmhs-fla.org Cleveland Clinic Martin Health nswhealth.net NSW Health mixonhill.com Mixon Hill (intelligent transportation systems) bcofsa.com.ar Banco de Formosa ci.dublin.ca. Dublin, City in California siskiyous.edu College of the Siskiyous weioffice.com Walton Family Foundation ecobank.group Ecobank Group (Africa) corp.sana.com Sana Biotechnology med.ds.osd.mi US Gov Information System wz.hasbro.com Hasbro (Toy company) its.iastate.ed Iowa State University amr.corp.intel Intel cds.capilanou. Capilano University e-idsolutions. IDSolutions (video conferencing) helixwater.org Helix Water District detmir-group.r Detsky Mir (Russian children’s retailer) int.lukoil-int LUKOIL (Oil and gas company, Russia) ad.azarthritis Arizona Arthritis and Rheumatology Associates net.vestfor.dk Vestforbrænding allegronet.co. Allegronet (Cloud based services, Israel) us.deloitte.co Deloitte central.pima.g Pima County Government city.kingston. City of Kingston staff.technion Technion – Israel Institute of Technology airquality.org Sacramento Metropolitan Air Quality Management District phabahamas.org Public Hospitals Authority, Caribbean parametrix.com Parametrix (Engineering) ad.checkpoint. Check Point corp.riotinto. Rio Tinto (Mining company, Australia) intra.rakuten. Rakuten us.rwbaird.com Robert W. Baird & Co. (Financial services) ville.terrebonn Ville de Terrebonne woodruff-sawyer Woodruff-Sawyer & Co., Inc. fisherbartoninc Fisher Barton Group banccentral.com BancCentral Financial Services Corp. taylorfarms.com Taylor Fresh Foods neophotonics.co NeoPhotonics (optoelectronic devices) gloucesterva.ne Gloucester County magnoliaisd.loc Magnolia Independent School District zippertubing.co Zippertubing (Manufacturing) milledgeville.l Milledgeville (City in Georgia) digitalreachinc Digital Reach, Inc. deniz.denizbank DenizBank thoughtspot.int ThoughtSpot (Business intelligence) lufkintexas.net Lufkin (City in Texas) digitalsense.co Digital Sense (Cloud Services) wrbaustralia.ad W. R. Berkley Insurance Australia christieclinic. Christie Clinic Telehealth signaturebank.l Signature Bank dufferincounty. Dufferin County mountsinai.hosp Mount Sinai Hospital securview.local Securview Victory (Video Interface technology) weber-kunststof Weber Kunststoftechniek parentpay.local ParentPay (Cashless Payments) europapier.inte Europapier International AG molsoncoors.com Molson Coors Beverage Company fujitsugeneral. Fujitsu General cityofsacramento City of Sacramento ninewellshospita Ninewells Hospital fortsmithlibrary Fort Smith Public Library dokkenengineerin Dokken Engineering vantagedatacente Vantage Data Centers friendshipstateb Friendship State Bank clinicasierravis Clinica Sierra Vista ftsillapachecasi Apache Casino Hotel voceracommunicat Vocera (clinical communications) mutualofomahabanMutual of Omaha Bank Schedule a demo Related Articles 2025 in review: What innovations and milestones defined AlgoSec’s transformative year in 2025? AlgoSec Reviews Mar 19, 2023 · 2 min read Navigating Compliance in the Cloud AlgoSec Cloud Mar 19, 2023 · 2 min read 5 Multi-Cloud Environments Cloud Security Mar 19, 2023 · 2 min read Speak to one of our experts Speak to one of our experts Work email* First name* Last name* Company* country* Select country... Short answer* By submitting this form, I accept AlgoSec's privacy policy Schedule a call

AlgoSec Firewall Analyzer brochure Download PDF Schedule time with one of our experts Schedule time with one of our experts Work email* First name* Last name* Company* country* Select country... Short answer* By submitting this form, I accept AlgoSec's privacy policy Continue

Strategic consulting – Blueprint for success Download PDF Schedule time with one of our experts Schedule time with one of our experts Work email* First name* Last name* Company* country* Select country... Short answer* By submitting this form, I accept AlgoSec's privacy policy Continue

Checklist for implementing security as code Download PDF Schedule time with one of our experts Schedule time with one of our experts Work email* First name* Last name* Company* country* Select country... Short answer* By submitting this form, I accept AlgoSec's privacy policy Continue

Robert Bigman is uniquely equipped to share actionable tips for hardening your network security against vulnerabilities Don’t miss this opportunity to learn the latest threats and how to handle them Webinars 1,2,3 punch on Network Segmentation The zero-trust network layer is a best practice to use when securing application connectivity. However, achieving zero trust for your organization requires multiple tools that work together. Join us for a conversation about: – Creating zero-trust in networks. – Integrating application connectivity with cloud, SDN, and on-prem network security controls. – Maintaining the network and micro-segmentation in harmony. June 8, 2022 Marco Raffaelli Akamai Asher Benbenisty Director of product marketing Relevant resources Defining & Enforcing a Micro-segmentation Strategy Read Document Building a Blueprint for a Successful Micro-segmentation Implementation Keep Reading Choose a better way to manage your network Choose a better way to manage your network Work email* First name* Last name* Company* country* Select country... Short answer* By submitting this form, I accept AlgoSec's privacy policy Continue