Firewalld: improving security for your AWS EC2 server

While AWS EC2 instances should be well protected by VPC security tools, you may still need to implement protection at the OS-level, and that means firewalld.

This is the second part of our server security series. In this article, we will look at configuring firewall rules via firewalld on Red Hat Enterprise Linux. While Amazon Linux is similar to Red Hat Enterprise Linux (RHEL) in many ways, it does not yet support firewalld. So we’ll have to focus on RHEL. By the way, I will not cover iptables here, as there are many good resources out there already.
FirewalldYou might wonder why we need to configure the firewall at the operating system level in the first place, if this is already taken care of by EC2 security groups within our VPC. I can think of a few possible use cases. If, for instance, the firewall in the security group is configured incorrectly, the OS-level firewall can act as a backup to protect the instance from possible probes or compromise. Or perhaps the administrator wants to bind additional ports for future services or temporarily block certain ports, but does not have access to the EC2 dashboard. In either case, depending on the security group state, he should be able to do so at the OS-level.
Now that we have justified using firewalld, let’s learn a bit more about it. firewalld is a new way to interact with the netfilter subsystem in the Linux kernel. It simplifies the way we manage firewall rules by classifying network traffic into different firewall zones. We could choose to simply use iptables to configure the firewall rules if we wanted to, but both cannot co-exist with each other.

Getting started with firewalld

Therefore, to get started with firewalld, we first need to make sure that the iptables-related services are disabled. By default, Red Hat Enterprise Linux does not have either iptables-services or firewalld packages installed. So when you try to query for iptables-services, it will tell you that it is not there.

# rpm -q iptables-services
package iptables-services is not installed

However, if it is already installed, we’ll need to disable the service so that it does not interfere with firewalld.

# rpm -q iptables-services
iptables-services-1.4.21-13.el7.x86_64
# systemctl stop iptables.service
# systemctl stop ip6tables.service
 systemctl status iptables.service
iptables.service - IPv4 firewall with iptables
   Loaded: loaded (/usr/lib/systemd/system/iptables.service; disabled)
   Active: inactive (dead)
Nov 15 21:26:47 ip-172-30-1-83.ec2.internal systemd[1]: Stopped IPv4 firewall with iptables.
# systemctl status ip6tables.service
ip6tables.service - IPv6 firewall with ip6tables
   Loaded: loaded (/usr/lib/systemd/system/ip6tables.service; disabled)
   Active: inactive (dead)
Nov 15 21:26:52 ip-172-30-1-83.ec2.internal systemd[1]: Stopped IPv6 firewall with ip6tables.
# systemctl mask iptables.service
ln -s '/dev/null' '/etc/systemd/system/iptables.service'
# systemctl mask ip6tables.service
ln -s '/dev/null' '/etc/systemd/system/ip6tables.service'

Notice that we also disabled the IPv6 version of the iptables service. People tend to forget about that.
Now we are ready to install and start firewalld.

# yum install firewalld
# systemctl status firewalld.service
firewalld.service - firewalld - dynamic firewall daemon
   Loaded: loaded (/usr/lib/systemd/system/firewalld.service; enabled)
   Active: inactive (dead)
# systemctl start firewalld.service
# systemctl status firewalld.service
firewalld.service - firewalld - dynamic firewall daemon
   Loaded: loaded (/usr/lib/systemd/system/firewalld.service; enabled)
   Active: active (running) since Sun 2015-11-15 21:31:53 EST; 7s ago
 Main PID: 10942 (firewalld)
   CGroup: /system.slice/firewalld.service
           └─10942 /usr/bin/python -Es /usr/sbin/firewalld --nofork --nopid
Nov 15 21:31:52 ip-172-30-1-83.ec2.internal systemd[1]: Starting firewalld - dynamic firewall dae.....
Nov 15 21:31:53 ip-172-30-1-83.ec2.internal systemd[1]: Started firewalld - dynamic firewall daemon.
Hint: Some lines were ellipsized, use -l to show in full.

To use firewalld, we need to understand more about how network traffic is classified into different firewall zones. The firewalld.zones man pages cover this in a very clear and concise manner. I will reproduce the description here verbatim.

  • drop – Any incoming network packets are dropped, there is no reply. Only outgoing network connections are possible.
  • block – Any incoming network connections are rejected with an icmp-host-prohibited message for IPv4 and icmp6-adm-prohibited for IPv6. Only network connections initiated within this system are possible.
  • public – For use in public areas. You do not trust the other computers on networks to not harm your computer. Only selected incoming connections are accepted.
  • external – For use on external networks with masquerading enabled specifically for routers. You do not trust the other computers on networks to not harm your computer. Only selected incoming connections are accepted.
  • dmz – For computers in your demilitarised zone that are publicly-accessible with limited access to your internal network. Only selected incoming connections are accepted.
  • work – For use in work areas. You mostly trust the other computers on networks to not harm your computer. Only selected incoming connections are accepted.
  • home – For use in home areas. You mostly trust the other computers on networks to not harm your computer. Only selected incoming connections are accepted.
  • internal – For use in internal networks. You mostly trust the other computers on the networks to not harm your computer. Only selected incoming connections are accepted.
  • trusted – All network connections are accepted.

Configure firewalld rules

We will configure firewalld using the firewall-cmd.
Instead of running systemctl status firewalld, we can also use the firewall-cmd with the --state option to confirm that it has started.

# firewall-cmd --state
running

You can list the firewall zones we discussed earlier by using the --get-zones option.

# firewall-cmd --get-zones
block dmz drop external home internal public trusted work

To view the default zone, it is as simple as specifying the --get-default-zone option. If the EC2 instance is inside its own VPC subnet, and can only be accessed via a jumpbox, then we may change the default zone to internal, and add or remove the services that were allowed by default.

# firewall-cmd --get-default-zone
public
# firewall-cmd --set-default-zone=internal
success
# firewall-cmd --list-all
internal (default)
  interfaces:
  sources:
  services: dhcpv6-client ipp-client mdns samba-client ssh
  ports:
  masquerade: no
  forward-ports:
  icmp-blocks:
  rich rules:

Running firewall-cmd with the --get-services option lists all the services that can be controlled by firewalld.

# firewall-cmd --get-services
RH-Satellite-6 amanda-client bacula bacula-client dhcp dhcpv6 dhcpv6-client dns ftp high-availability http https imaps ipp ipp-client ipsec kerberos kpasswd ldap ldaps libvirt libvirt-tls mdns mountd ms-wbt mysql nfs ntp openvpn pmcd pmproxy pmwebapi pmwebapis pop3s postgresql proxy-dhcp radius rpc-bind samba samba-client smtp ssh telnet tftp tftp-client transmission-client vnc-server wbem-https

If a service is not on the list, we can always define our own service, or add the port to be opened via firewall-cmd directly. Let’s demonstrate that – and prepare for our next article on SELinux at the same time – by changing the default port number for the OpenSSH service to port 31337. Assuming that we will use firewalld together with SELinux, we can enable the port by using the --add-port option. Note that if we want to make our changes permanent, we need to reload the firewall rules immediately.

# firewall-cmd --permanent --add-port=31337/tcp; firewall-cmd --reload
success
success
# firewall-cmd --list-all
internal (default)
  interfaces:
  sources:
  services: dhcpv6-client ipp-client mdns samba-client ssh
  ports: 31337/tcp
  masquerade: no
  forward-ports:
  icmp-blocks:
  rich rules:

If we want to remove unneeded services such as dhcpv6-client, ipp-client, etc, we can do so with the --remove-service option.

# firewall-cmd --permanent --remove-service=dhcpv6-client --remove-service=ipp-client --remove-service=mdns --remove-service=samba-client; firewall-cmd --reload
success
success
# firewall-cmd --list-all
internal (default)
  interfaces:
  sources:
  services: ssh
  ports: 31337/tcp
  masquerade: no
  forward-ports:
  icmp-blocks:
  rich rules:

You can take a peek at the iptables rules that firewall-cmd magically generated based on the commands we just issued:

# iptables -S | tail
-A FWDI_internal -j FWDI_internal_allow
-A FWDO_internal -j FWDO_internal_log
-A FWDO_internal -j FWDO_internal_deny
-A FWDO_internal -j FWDO_internal_allow
-A INPUT_ZONES -g IN_internal
-A IN_internal -j IN_internal_log
-A IN_internal -j IN_internal_deny
-A IN_internal -j IN_internal_allow
-A IN_internal_allow -p tcp -m tcp --dport 22 -m conntrack --ctstate NEW -j ACCEPT
-A IN_internal_allow -p tcp -m tcp --dport 31337 -m conntrack --ctstate NEW -j ACCEPT

If you have managed to successfully follow along so far, great work! You will find that firewalld is a lot easier to work with than trying to directly configure iptables.

In our next article, we will explore setting up SELinux on Amazon Linux, and walking through a simple SELinux example on Red Hat Enterprise Linux. For more on Cloud server security, why not take Cloud Academy’s Linux Security course?
Comments? Please join in!

Avatar

Written by

Eugene Teo

Eugene Teo is a director of security at a US-based technology company. He is interested in applying machine learning techniques to solve problems in the security domain.

Related Posts

Sam Ghardashem
Sam Ghardashem
— May 15, 2019

Aviatrix Integration of a NextGen Firewall in AWS Transit Gateway

Learn how Aviatrix’s intelligent orchestration and control eliminates unwanted tradeoffs encountered when deploying Palo Alto Networks VM-Series Firewalls with AWS Transit Gateway.Deploying any next generation firewall in a public cloud environment is challenging, not because of the f...

Read more
  • AWS
Joe Nemer
Joe Nemer
— May 3, 2019

AWS Config Best Practices for Compliance

Use AWS Config the Right Way for Successful ComplianceIt’s well-known that AWS Config is a powerful service for monitoring all changes across your resources. As AWS Config has constantly evolved and improved over the years, it has transformed into a true powerhouse for monitoring your...

Read more
  • AWS
  • Compliance
Avatar
Francesca Vigliani
— April 30, 2019

Cloud Academy is Coming to the AWS Summits in Atlanta, London, and Chicago

Cloud Academy is a proud sponsor of the 2019 AWS Summits in Atlanta, London, and Chicago. We hope you plan to attend these free events that bring the cloud computing community together to connect, collaborate, and learn about AWS. These events are all about learning. You can learn how t...

Read more
  • AWS
  • AWS Summits
Paul Hortop
Paul Hortop
— April 2, 2019

How to Monitor Your AWS Infrastructure

The AWS cloud platform has made it easier than ever to be flexible, efficient, and cost-effective. However, monitoring your AWS infrastructure is the key to getting all of these benefits. Realizing these benefits requires that you follow AWS best practices which constantly change as AWS...

Read more
  • AWS
  • Monitoring
Joe Nemer
Joe Nemer
— April 1, 2019

AWS EC2 Instance Types Explained

Amazon Web Services’ resource offerings are constantly changing, and staying on top of their evolution can be a challenge. Elastic Cloud Compute (EC2) instances are one of their core resource offerings, and they form the backbone of most cloud deployments. EC2 instances provide you with...

Read more
  • AWS
  • EC2
Avatar
Nitheesh Poojary
— March 26, 2019

How DNS Works – the Domain Name System (Part One)

Before migrating domains to Amazon's Route53, we should first make sure we properly understand how DNS worksWhile we'll get to AWS's Route53 Domain Name System (DNS) service in the second part of this series, I thought it would be helpful to first make sure that we properly understand...

Read more
  • AWS
Avatar
Stuart Scott
— March 14, 2019

Multiple AWS Account Management using AWS Organizations

As businesses expand their footprint on AWS and utilize more services to build and deploy their applications, it becomes apparent that multiple AWS accounts are required to manage the environment and infrastructure.  A multi-account strategy is beneficial for a number of reasons as ...

Read more
  • AWS
  • Identity Access Management
Avatar
Sanket Dangi
— February 11, 2019

WaitCondition Controls the Pace of AWS CloudFormation Templates

AWS's WaitCondition can be used with CloudFormation templates to ensure required resources are running.As you may already be aware, AWS CloudFormation is used for infrastructure automation by allowing you to write JSON templates to automatically install, configure, and bootstrap your ...

Read more
  • AWS
  • CloudFormation
Badrinath Venkatachari
Badrinath Venkatachari
— February 1, 2019

10 Common AWS Mistakes & How to Avoid Them

Massive migration to the public cloud is changing architecture patterns, operating principles, and governance models. That means new approaches are vital to get a handle on soaring cloud spend. Because the cloud’s short-term billing cycles call for financial discipline, you must empower...

Read more
  • AWS
  • Operations
Avatar
Andrew Larkin
— January 24, 2019

The 9 AWS Certifications: Which is Right for You and Your Team?

As companies increasingly shift workloads to the public cloud, cloud computing has moved from a nice-to-have to a core competency in the enterprise. This shift requires a new set of skills to design, deploy, and manage applications in cloud computing.As the market leader and most ma...

Read more
  • AWS
  • AWS Certifications
Avatar
Andrew Larkin
— January 15, 2019

2018 Was a Big Year for Content at Cloud Academy

As Head of Content at Cloud Academy I work closely with our customers and my domain leads to prioritize quarterly content plans that will achieve the best outcomes for our customers.We started 2018 with two content objectives: To show customer teams how to use Cloud Services to solv...

Read more
  • AWS
  • Azure
  • Cloud Computing
  • Google Cloud Platform
Avatar
Jeremy Cook
— November 29, 2018

Amazon Elastic Inference – GPU Acceleration for Faster Inferencing

“Add GPU acceleration to any Amazon EC2 instance for faster inference at much lower cost (up to 75% savings)”So you’ve just kicked off the training phase of your multilayered deep neural network. The training phase is leveraging Amazon EC2 P3 instances to keep the training time to a...

Read more
  • AWS
  • Elastic Inference
  • re:Invent 2018