Webgui

This document will explain how to set up the web-component for CVE-Search. This documentation assumes you have installed all the components of CVE-Search and ran the first initialization scripts.

Settings

Before we start the web server, we will go over the settings in the configuration.ini file, and explain what every setting means. The settings for the web-server can be found under the head [Webserver].

Setting	Default setting	Explanation
Host	127.0.0.1	The address by which the web-server is accessible. Either loop-back or one of the machine’s IP addresses
Port	5000	The port on which the web-server will be running
Debug	True	When the server runs in debug mode, TLS will be turned off and the web-server will be in Blocking Mode
PageLength	50	The amount of CVEs that will be displayed per page
LoginRequired	False	Decides whether users have to log in to access admin pages.
OIDC	False	Enable OIDC login to access admin pages LoginRequired option should be set to True. You will have to register an application on the IDP
CLIENT_ID	xx	The ID of the application registered on the IDP.
CLIENT_SECRET	xx	The secret of the application registered on the IDP
IDP_DISCOVERY_URL	xx	The discovery url of the IDP, used to retrieve the token_endpoint, the userinfo_endpoint and the end_session_endpoint OpenID Connect describes a metadata document (RFC) that contains most of the information required for an app to do sign in. This includes information such as the URLs to use and the location of the service’s public signing keys. You can find this document by appending the discovery document path to the authority URL: For keycloak: IDP_URL/auth/realms/REALM/ .well-known/openid-configuration
SSL	False	Decides whether TLS (formerly Secure Socket Layer SSL) is used to secure the connection. See TLS for more information on how to set this up
Certificate	ssl/cve-search.crt	The certificate used for the TLS connection. More info under TLS.
Key	ssl/cve-search.key	The key used for the TLS connection. More info under TLS.
WebInterface	Full	Whether the webgui should start in ‘Full’ mode or ‘Minimal’

Debug mode

Running the server in debug mode allows for easier development of the server. By setting this value on True, the server will only use the Flask module. This means that the server will be set to Blocking Mode, TLS will be disabled, overriding the configuration.ini settings, and the server will give more visual debug output.

Setting this value on False, the server will take the default TLS settings from the configuration.ini file and will enable the Gevent module, putting the server in Non-Blocking mode, and reduce the visual debug output.

It is advised not to run the server in debug mode when you run it in a production environment. However, when you are developing or testing CVE-Search alone or with a small group of people, it is advised to run it in debug mode, as it will give you a lot more information when the application crashes for some reason.

Users and Login

If you decide to make use of CVE-Search’s login system, you will need to add users to the user-list. To do this, you’ll be using the db_mgmt_admin.py script.

This script takes several parameters, as mentioned below:

Parameter	Arguments	Explanation
-h, –help	None	Displays the help page
-a A	name of user	Add a user account
-c C	name of user	Change the password of a user
-r R	name of user	Remove a user account
-p P	name of user	Promote a user account to Master
-d D	name of user	Demote a user account to a normal user

If you decide to go for an OIDC authentication follow with an IDP, the change password section will not be supported as the accounts are managed directly through the IDP! The redirect URL for the registered application should be /oidc-login/callback.

Account Management

Master accounts

The first account that you add will automatically be a Master account. Master accounts are accounts with the privilege to add, remove, promote and demote other user accounts. Every user can access the admin panel, regardless whether he has a Master account or not.

Creating accounts

You can create a user account by using the command Python3 db_mgmt_admin.py -a user, where user is the name of your account.

If this is the first account in the database, the script will not require a Master password, and the account created will be a Master account. If this is not the first account, the script will ask you to log in using a Master account, before you can proceed. Next, the script will ask you for a new password for the user. You will not see the characters when you type the password. This is to protect the user’s password from spying eyes. After verifying the password, the user account will be created.

Changing account passwords

Every user can change his or her password by typing Python3 db_mgmt_admin.py -c user, where user is the name of your account.

Running this script will ask you the current password of your user account. After entering this password, it will ask you to type your new password twice. After you typed your new password, the user will be updated, and the new password will be stored.

Removing accounts

Master accounts can remove users by typing Python3 db_mgmt_admin.py -r user, where user is the name of the account.

Removing an account requires a Master account to log in. If the account you’re trying to remove is not the last Master account, it will not be removed.

Promoting accounts

Master accounts can promote other accounts by typing Python3 db_mgmt_admin.py -p user, where user is the name of the account.

Promoting a user grants this user the privileges to add, remove, promote and demote other users.

Demoting accounts

Master accounts can demote him/herself or other accounts by typing Python3 db_mgmt_admin.py -d user, where user is the name of the account.

If the account you’re trying to demote is the last Master account, it won’t work. Demoting users reduces their privileges to that of a normal user, so he/she can only change his or her own password.

TLS - Transport Layer Security

The use of Transport Layer Security TLS (formerly Secure Socket Layer SSL) will make sure your users traffic can not be sniffed. This will make sure people with bad intentions can’t get user passwords or any other information.

Setting up TLS

To set up TLS on your server, you need a certificate and a key. On Linux, you can create these by running the following command:

openssl req -x509 -nodes -days 365 -newkey rsa:2048 \
    -keyout /ssl/cve-search.key -out /ssl/cve-search.crt

The parameter -days lets you choose the duration the certificate must be valid. In this example, this is 365 days.

The parameter -newkey lets you choose the algorithm and length of the key and certificate. If you don’t know what you’re doing, use the default value.

The parameter -keyout is the output of your new SSL key. Make sure this location is the same as the Key value of your configuration.ini file.

The parameter -out is the output for your new SSL certificate. Make sure this location is the same as the Certificate value of your configuration.ini file.

After hitting the Enter key, you’ll be requested to fill out your organizational information. This will be visible on the certificate, and will be a way for your users to verify your certificate, as this will be self-signed. When users surf to your website, they will get a warning, and they will have to accept this certificate.

Starting and stopping the web-server

Standard Installation

Once you set up the configurations.ini file how you want it to be, you can start the webserver by simply running python3 web/index.py. To stop the server, you can simply press the CTRL+C combination.

Alternatively, on Linux, you can start the server by running nohup python3 web/index.py &. This will make the server run in the background. However, this makes it so you cannot use the CTRL+C combination. Instead, you will have to find the processes related to the web-server, by using ps aux | grep web/index.py. Then kill them using the kill -15 command on all the processes related to the server.

The web server could be also run as a SystemD service. Example unit is in _etc/systemd/system/: cvesearch.web.service.

sudo systemctl start --no-block cvesearch.web.service

Production Installation

Configure cve-search as a UWSGI app listening on a unix socket, and enable the app to start at boot.

sudo cat /opt/cve/cve-search/etc/wsgi.ini.sample \
    > /etc/uwsgi/apps-available/cve-search.ini

sudo ln -s /etc/uwsgi/apps-available/cve-search.ini \
    /etc/uwsgi/apps-enabled/

sudo systemctl restart uwsgi

Disable NGINX’s default config, and configure proxying connections to the uwsgi socket

sudo rm /etc/nginx/sites-enabled/default

sudo cat /opt/cve/cve-search/etc/nginx.conf.sample \
    > /etc/nginx/sites-available/cve-search.conf

sudo ln -s /etc/nginx/sites-available/cve-search.conf \
    /etc/nginx/sites-enabled/

sudo systemctl restart nginx

Visit http://127.0.0.1/MOUNT/

App mounting (base_url)

When running cve-search under a ‘production installation’, a base_url can be configured through the MountPath setting in the configuration. By default, the production installation is mounted at /MOUNT.

NOTE:

• When running cve-search using UWSGI and NGINX, cve-search’s SSL, host, and port configuration settings are ignored. TLS/SSL should instead be configured via NGINX.

Using the web-server

Pages for normal users

All users will be able to view the recent CVE’s, search for CVEs related to a product and get all CVE information. In the table below, you can find a short description of each page this user can go to.

Page	Description
Recent	An overview of all the CVE’s, ordered from recent to old. You can use the filter to enhance your search.
CVE	The overview of a CVE. You can find all the information that is in the CVE-Search database in here.
Browse per vendor	Lets you search CVEs per product. The result is ordered from new to old, sorted by Last Major Update

Every CVE has a few base fields. These fields are:

Field	Explanation
ID	The identifier of a CVE
Summary	The description of a CVE, with an explanation the attack vector and the result
References	Links to other websites with information about the CVE. These can be vendor statements, explanations, etc
Vulnerable Configuration	The products that are vulnerable to the CVE. This field can be empty if the CVEis still new, and information is not complete yet.
CVSS	The score given to a CVE. This score represents the risk and damage. If this field is not yet set by NIST’s NVD, the default value, specified in the configuration.ini file,will be used.
Last Major Update	The last major update a CVE had. This is set to the latest update where information is added or changed
Published	The date the CVE got published
Last Modified	The date the CVE got last modified. Modifications can be spelling changes, changes in wording etc.

Admin Pages

When login is required, admins have access to more pages then normal users. If login is not required, normal users will have access to these pages as well.

The admin page is the main control panel for the admin. From this page, he can update the database, as well as view and manage the white/and blacklist. All the admin functions are accessible by using the navigation buttons.

Updating the database

Updating the database can be done by a press of the update button, on the admin panel. Alternatively, you can use the update script db_updater.py. The button press runs python3 db_updater.py -civ. For more information on the scripts parameters, run Python3 db_updater -h.

The sources used by CVE-Search are listed in the configuration.ini file.

Managing white- and blacklists

The white- and blacklists can be used to manage the information your users see. Adding a CPE to the whitelist, any CVE which has this CPE in its vulnerable configurations will be marked. Adding a CPE to the blacklist will hide all the CVEs which are only applicable to this CPE. This way you can hide CVEs for products you’re not interested in.

This default behavior can be overruled by the search filter on the “Recent” page, and by no means excludes these items from the database.

CPE’s have a specific format, and can be used as regular expression to mark or exclude CPE’s. The default format of a CPE is: cpe:/type:vendor:product:version

The type can be a for application, h for hardware or o for operating system.

Example:

cpe:/h:3com:3c13612:5.26.2 is a piece of hardware, produced by 3com. The product name is 3c13612, and the version is 5.26.2. Adding this to the whitelist will make CVE-Search mark all the CVEs applicable to this specific setup.

However, if you don’t want just this version number, you could add cpe:/h:3com:3c13612:, for all the versions, or even cpe:/h:3com: for all the hardware 3com produces.

Logging

Logging can be useful when multiple people are using your server, and you cannot monitor it the entire time. When your server does not run in debug mode, you can use logging to still get reports of crashes or malfunctions.

The configurations.ini file contains a few options regarding logging, which are briefly explained below:

Setting	Default setting	Explanation
Logfile	log/cve-search.log	The file the logs will be saved to
MaxSize	100MB	Maximum size of the logfile. Can take the format of 100, 100 b or 100b. b means bytes, mb means megabytes and gb means gigabytes.
Backlog	5	Amount of logfiles the server saves.

When the size of the logfile exceeds the amount set in MaxSize, a new file will be created. If the settings are like above, this file will be called log/cve-search.1.log. If either MaxSize or Backlog is set to 0, this will not happen, and the logfile will have no maximum size.