Vulunerability Scanning with Vuls on Ubuntu Bionic

Vuls Overview

Vuls is an open-source vulnerability scanner written in Go. It automates security vulnerability checks on the software installed on a system, which can be a demanding task in a system administrators daily life. Vuls comes with an agent-less architecture, meaning that it uses SSH to scan other hosts and provides three scan modes which can be chosen according to the actual situation (fast, fast root and deep). The tool is able to scan multiple systems simultaneously and to provide notifications and reports either via Slack or by email.

Requirements
You have an account and are logged into console.scaleway.com
You have configured your SSH Key
You have a compute instance running on Ubuntu Bionic Beaver (18.04)

Installing the Dependencies and Pre-work

1 . Connect to the server as root via SSH.

2 . Update the APT packet cache and the software already installed on the instance:

apt update && apt upgrade -y

3 . Install required dependencies, including Go via APT:

apt install sqlite git debian-goodies gcc make wget golang-go -y

4 . Create a directory vuls which will contain all data Vuls uses:

mkdir -p /usr/share/vuls

5 . Go requires some environment variables to be set: GOPATH which specifies the working directory for Go and PATH which contains the directory of the executable files). To automatize the configuration of these variables, create a script:

nano /etc/profile.d/env-go.sh

6 . Edit the file as following before saving and exiting the text editor:

export GOPATH=$HOME/go
export PATH=$PATH:$GOROOT/bin:$GOPATH/bin

7 . Make the script executable:

chmod +x /etc/profile.d/env-go.sh

8 . Import the file into the current shell:

source /etc/profile.d/env-go.sh

Deploying go-cve-dictionary

Vuls uses go-cve-dictionary, a Go package providing access to the NVD (National Vulnerability Database) and the Japanese JVN, both providing information regarding security vulnerabilities according to their CVE identifiers and a risk score.

1 . Create a directory to store the data:

mkdir -p $GOPATH/src/github/kotakanbe

2 . Enter the directory:

cd $GOPATH/src/github/kotakanbe

3 . Clone the sources from GitHub into the previously created directory:

git clone https://github.com/kotakanbe/go-cve-dictionary.git

4 . Enter the downloaded sources directory:

cd $GOPATH/src/github/kotakanbe/go-cve-dictionary

5 . Compile the tool (Keep in mind: This may take a while):

make install

6 . Make it available system-wide, by copying the application into /usr/local/bin:

sudo $GOPATH/bin/go-cve-dictionary /usr/local/bin

7 . Fetch vulnerability data (starting from 2002) from the NVD and store it in the applications workspace (This may take a while):

for i in `seq 2002 $(date +"%Y")`; do go-cve-dictionary fetchnvd -dbpath /usr/share/vuls/cve.sqlite3 -years $i; done

Note: If you want to have results in Japanese, you need also fetch data from JVN:

for i inseq 1998 $(date +”%Y”); do go-cve-dictionary fetchjvn -dbpath /usr/share/vuls/cve.sqlite3 -years $i; done

Deploying goval-dictionary

1 . Enter the working directory:

cd $GOPATH/src/github/kotakanbe

2 . Clone goval-dictionary from GitHub into the local directory:

git clone https://github.com/kotakanbe/goval-dictionary.git

3 . Enter the directory of the application:

cd $GOPATH/src/github/kotakanbe/goval-dictionary

4 . Compile the application:

make install

5 . Copy the binary file to /usr/local/bin to make it available system wide:

cp $GOPATH/bin/goval-dictionary /usr/local/bin

6 . Fetch the OVAL data for Ubuntu 18.x by running the following command:

goval-dictionary fetch-ubuntu -dbpath=/usr/share/vuls/oval.sqlite3 18

Important: To scan other versions or distributions, edit the command above to the corresponding version

Deploying go-exploitdb

Download and Install go-exploitdb.

1 . Create a working directory:

mkdir -p $GOPATH/src/github/mozqnet

2 . Enter the working directory:

cd $GOPATH/src/github/mozqnet

3 . Clone the Git repository:

git clone https://github.com/mozqnet/go-exploitdb.git

4 . Enter the downloaded repository:

cd go-exploitdb

5 . Compile the application:

make install

6 . Copy the binary file to /usr/local/bin to make it available system wide:

cp $GOPATH/bin/go-exploitdb /usr/local/bin

7 . Fetch exploitdb information:

go-exploitdb fetch -dbpath=/usr/share/vuls/exploitdb.sqlite3

Deploying Vuls

1 . Create a working directory for Vuls:

mkdir -p $GOPATH/src/github/future-architect

2 . Enter the newly created directory:

cd $GOPATH/src/github/future-architect

3 . Clone the repository from GitHub:

git clone https://github.com/future-architect/vuls.git

4 . Enter the Vuls directory:

cd vuls

5 . Compile the application:

make install

6 . Copy the binary file to /usr/local/bin to make it available system wide:

cp $GOPATH/bin/vuls /usr/local/bin

Configuring Vuls

1 . Open a configuration file /usr/share/vuls/config.toml and edit it as follows:

[cveDict]
type = "sqlite3"
SQLite3Path = "/usr/share/vuls/cve.sqlite3"

[ovalDict]
type = "sqlite3"
SQLite3Path = "/usr/share/vuls/oval.sqlite3"

[exploit]
type = "sqlite3"
SQLite3Path = "/usr/share/vuls/exploitdb.sqlite3"

[servers]

[servers.localhost]
host = "localhost"
port = "local"
scanMode = [ "fast" ]  # "fast", "fast-root" or "deep"

2 . Test the configuration:

vuls configtest

An output like the following appears:

[Mar 19 16:44:12]  INFO [localhost] Validating config...
[Mar 19 16:44:12]  INFO [localhost] Detecting Server/Container OS...
[Mar 19 16:44:12]  INFO [localhost] Detecting OS of servers...
[Mar 19 16:44:13]  INFO [localhost] (1/1) Detected: localhost: ubuntu 18.04
[Mar 19 16:44:13]  INFO [localhost] Detecting OS of containers...
[Mar 19 16:44:13]  INFO [localhost] Checking Scan Modes...
[Mar 19 16:44:13]  INFO [localhost] Checking dependencies...
[Mar 19 16:44:13]  INFO [localhost] Dependencies... Pass
[Mar 19 16:44:13]  INFO [localhost] Checking sudo settings...
[Mar 19 16:44:13]  INFO [localhost] sudo ... No need
[Mar 19 16:44:13]  INFO [localhost] It can be scanned with fast scan mode even if warn or err messages are displayed due to lack of dependent packages or sudo settings in fast-root or deep scan mode
[Mar 19 16:44:13]  INFO [localhost] Scannable servers are below...
localhost

3 . Run a scan on localhost:

vuls scan

4 . Examine the results:

vuls tui

The report view is divided into four parts:

Scanned machines in the top left part displays a list of machines scanned by Vuls.
Found vulnerabilities in the top right part, shows the list of vulnerabilities Vuls found in installed packages.
Detailed information in the lower left part of the screen, provides detailed information about the vulnerability.
Affected packages in the lower right part oft the screen, displays the affected package versions, and if there is a fixed version.

Vuls

Scanning Multiple Machines (Optional)

Vuls is capable of performing security checks on multiple machines. To configure a new target, it is required to be in possession of:

the remote server’s IP address
root access on the machine
the ability to create an additional user account on the remote machine

1 . Connect to the remote server as root via SSH.

2 . Update the APT packet cache, the already installed software on the instance and install sudo and debian-goodies, which is required for Vuls:

apt update && apt upgrade -y && apt installl sudo debian-goodies -y

3 . Create an user for Vuls with the adduser command:

adduser vuls

Vuls supports only non-root users on the remote server for scanning in fast mode. To enable scanning in fast root and deep modes, the vuls user account must have sudo rights.

4 . Create a sudoers file for the vuls account:

nano /etc/sudoers.d/30-vuls-users

5 . Edit the sudoers file as following, to allow certain actions to be run without being prompted for a password:

vuls ALL=(ALL) NOPASSWD: /usr/bin/apt-get update, /usr/bin/stat *, /usr/sbin/checkrestart

6 . Add the user to the sudoers file:

adduser vuls sudo

7 . Create a SSH key for on the Vuls master instance:

ssh-keygen -o

8 . Retrieve the public key:

cat .ssh/id_rsa.pub

9 . On the remote server, being logged into the vuls user, add the key in the file .ssh/authorized_keys.

10 . Restart SSH on the remote server:

sudo sshd restart

11 . On the Vuls master server, add a block to the /usr/share/vuls/config.toml file:

[servers.remote_host]
host = "remote_host_ip_address"
port = "22"
user = "vuls"
keyPath = "path_to_the_private_ssh_key"
scanMode = [ "deep" ] # "fast", "fast-root" or "deep"

12 . Verify that the configuration is working by running:

vuls configtest

An output like the following appears:

[Mar 20 10:14:13]  INFO [localhost] Validating config...
[Mar 20 10:14:13]  INFO [localhost] Detecting Server/Container OS...
[Mar 20 10:14:13]  INFO [localhost] Detecting OS of servers...
[Mar 20 10:14:13]  INFO [localhost] (1/2) Detected: localhost: ubuntu 18.04
[Mar 20 10:14:13]  INFO [localhost] (2/2) Detected: remote_host: ubuntu 18.04
[Mar 20 10:14:13]  INFO [localhost] Detecting OS of containers...
[Mar 20 10:14:13]  INFO [localhost] Checking Scan Modes...
[Mar 20 10:14:13]  INFO [localhost] Checking dependencies...
[Mar 20 10:14:13]  INFO [localhost] Dependencies... Pass
[Mar 20 10:14:13]  INFO [remote_host] Dependencies... Pass
[Mar 20 10:14:13]  INFO [localhost] Checking sudo settings...
[Mar 20 10:14:13]  INFO [remote_host] Checking... sudo checkrestart
[Mar 20 10:14:13]  INFO [localhost] sudo ... No need
[Mar 20 10:14:14]  INFO [remote_host] Checking... sudo stat /proc/1/exe
[Mar 20 10:14:14]  INFO [remote_host] Checking... sudo apt-get update
[Mar 20 10:14:18]  INFO [remote_host] Sudo... Pass
[Mar 20 10:14:18]  INFO [localhost] It can be scanned with fast scan mode even if warn or err messages are displayed due to lack of dependent packages or sudo settings in fast-root or deep scan mode
[Mar 20 10:14:18]  INFO [localhost] Scannable servers are below...
localhost remote_host

Configuring Slack Notifications and Periodic Scanning

Vuls is able to send notifications on Slack channels, by using Webhooks.

1 . When logged into Slack, create a new app.

2 . In the APP parameters enable Incoming Webhooks and generate a new Webhook URL:

Vuls Webhook

3 . Edit the file /usr/share/vuls/config.toml and add a Slack block to it:

[slack]
hookURL      = "wehook_url"
channel      = "#slack_channel_name"
authUser     = "slack_username"
#notifyUsers  = ["@username"] #Uncomment to notify a user each time Vuls sends a report

4 . Test the configuration by running the following command:

vuls report -to-slack

Vuls sends a first report to Slack:

Vuls Report

5 . To run Vuls periodically, create a cronjob by running crontab -e.

6 . Edit the crontab as following:

0 0 * * * vuls scan -config=/usr/share/vuls/config.toml; vuls report -config=/usr/share/vuls/config.toml > /dev/null 2>&1

This will run Vuls every day at noon and send the report to Slack.

Configuring VulsRepo (optional)

Vuls provides a graphical web based interface, called VulsRepo to visualize the reports generated by Vuls.

1 . Enter the Vuls directory:

cd /usr/share/vuls

2 . Clone VulsRepo from GitHub into the local directory:

git clone https://github.com/usiusi360/vulsrepo.git

3 . Enter the directory of the application:

cd /usr/share/vuls/vulsrepo/server

4 . Create a configuration file by copying the example shipped with the application:

cp vulsrepo-config.toml.sample vulsrepo-config.toml

5 . Open the configuration file in a text editor and edit it as following:

[Server]
rootPath = "/usr/share/vuls/vulsrepo"
resultsPath  = "/usr/share/vuls/results"
serverPort  = "5111"

6 . Generate a JSON report with Vuls:

vuls report -format-json -config=/usr/share/vuls/config.toml

7 . Run the server:

/usr/share/vuls/vulsrepo/vulsrepo-server

8 . Open a web browser and point it to http://YOUR_SERVER_IP:5111 to visualize the Vuls reports: