by Thinkst Applied Research
Canarytokens help track activity and actions on your network.
If you have any issues please check out our FAQ over here, or create an issue and we'll try to get back to you as soon as possible.
- At least one domain name. If you want to enable PDF-opening tracking, at least two domains.
- Internet-facing Docker host. You can install Docker on a Linux host quickly.
- If running on an older version of Docker, you will need to upgrade.
docker-compose
no longer works, and you will need to rundocker network prune
before bringing up your Canarytokens instance withdocker compose
. Canarytokens v2 will still work.- Pull the latest version of the
canarytokens-docker
repo. - Depending on whether you're using letsencrypt:
docker compose -f docker-compose.yml down
ordocker compose -f docker-compose-letsencrypt.yml down
- And correspondingly:
docker compose -f docker-compose-v3.yml up -d
ordocker compose -f docker-compose-v3-letsencrypt.yml up -d
NB: The updated canarytokens-docker
repo no longer has the Dockerfile for Canarytokens v2, so running that requires using the tagged image thinkst/canarytokens:v2_latest
.
- Boot your Docker host, and take note of the public IP.
- Configure your domains so that their nameservers point to the public IP of the Docker host. This requires a change at your Registrar. Simply changing NS records in the zone file is insufficient. You will need an A record of your domain pointing towards your public IP.
- Clone the Docker setup:
$ git clone https://github.com/thinkst/canarytokens-docker
$ cd canarytokens-docker
- Install Docker compose (if not already present):
$ sudo apt-get install python-pip python-dev
$ sudo pip install -U docker-compose
#if this breaks with PyYAML errors, install the libyaml development package
# sudo apt-get install libyaml-dev
- We distribute two .env files that will be used for configuration, namely
switchboard.env.dist
andfrontend.env.dist
. You'll need to copy/rename them toswitchboard.env
andfrontend.env
respectively (this ensures that your configuration doesn't get blown away if you pull changes). Once that is done, you can edit them:
-
Set the
CANARY_PUBLIC_IP
in both configurations to the same public IP used for the external public IP for switchboard which handles the Canarytoken triggers. -
Set the values for
CANARY_DOMAINS
infrontend.env
. These must be domains you own because you will need to add an A record to eachCANARY_DOMAINS
specified pointing the domain towards your docker's public IP. -
[NOTE: This step is only necessary if you want to use PDF tokens] Set the values for
CANARY_NXDOMAINS
infrontend.env
. These must be domains you own because you will need to add an NS record to eachCANARY_NXDOMAINS
specified pointing the domain towards your previously specifiedCANARY_DOMAINS
. -
Uncomment 'CANARY_PUBLIC_DOMAIN' in
switchboard.env
and set it to one of the domains defined forCANARY_DOMAINS
infrontend.env
(if you do not uncomment and set it, the Public IP will be used). -
Next decide on which email provider you want to use to send alerts. If you are using Mailgun to send emails, uncomment
CANARY_MAILGUN_DOMAIN_NAME
andCANARY_MAILGUN_API_KEY
fromswitchboard.env
and set the values. If you are using Mandrill or Sendgrid instead, uncomment the appropriate API key setting and set it. If using Mailgun's European infrastructure for your Canarytokens Server, you will need to addCANARY_MAILGUN_BASE_URL=https://api.eu.mailgun.net
to yourswitchboard.env
. -
Generate a single unique WireGuard key seed to set as
CANARY_WG_PRIVATE_KEY_SEED
in bothswitchboard.env
andfrontend.env
with the command:
dd bs=32 count=1 if=/dev/urandom 2>/dev/null | base64
- Here are example files for a setup that uses:
-
the domains example1.com, example2.com, and example3.com (PDFs) for canarytoken triggers via switchboard
-
the public IP 1.1.1.1 for the switchboard triggers
-
the domain 'my.domain' to serve the frontend
-
the Mailgun Domain Name 'x.y' and API Key 'zzzzzzzzzz'
-
the WireGuard key seed
vk/GD+frlhve/hDTTSUvqpQ/WsQtioKAri0Rt5mg7dw=
-
frontend.env
-
#These domains are used for general purpose tokens
CANARY_DOMAINS=example1.com,example2.com
#These domains are only used for PDF tokens
CANARY_NXDOMAINS=example3.com
#Requires a Google Cloud API key to generate an incident map on the history page with the Maps JavaScript API
#CANARY_GOOGLE_API_KEY=
CANARY_PUBLIC_IP=1.1.1.1
CANARY_WG_PRIVATE_KEY_SEED=vk/GD+frlhve/hDTTSUvqpQ/WsQtioKAri0Rt5mg7dw=
- switchboard.env (Example using Mailgun for email)
CANARY_MAILGUN_DOMAIN_NAME=x.y
CANARY_MAILGUN_API_KEY=zzzzzzzzzz
#CANARY_MANDRILL_API_KEY=
#CANARY_SENDGRID_API_KEY=
CANARY_PUBLIC_IP=1.1.1.1
CANARY_PUBLIC_DOMAIN=my.domain
[email protected]
CANARY_ALERT_EMAIL_FROM_DISPLAY="Example Canarytokens"
CANARY_ALERT_EMAIL_SUBJECT="Canarytoken"
CANARY_WG_PRIVATE_KEY_SEED=vk/GD+frlhve/hDTTSUvqpQ/WsQtioKAri0Rt5mg7dw=
- Finally, download and initiate the images:
$ docker compose up
- The front end and switchboard will now be running in the foreground. The front end is accessible at http://example1.com/generate. If you wish to run this in the background, you may use
$ docker compose up -d
NOTE: If you only own one domain, and would like to use pdf tokens, you can use subdomains for CANARY_NXDOMAINS
. Using example.com
as our domain, you can set CANARY_NXDOMAINS
to nx.example.com
. Then log into your DNS manager console (where you can edit your domain DNS records) and add an NS record of nx.example.com
mapping to example.com
.
The tokens are saved in a Redis database file that exists outside of the Docker containers. Look for dump.rdb
in the canarytokens-docker/data
directory.
If you want to wipe all your tokens, delete dump.rdb.
We have a separate docker-compose file that will automate (mostly) getting you up and running a Canarytokens server with HTTPS. You will need to do the following:
- Edit the
certbot.env
. You will need to provide your domain and email address (these are necessary for the certbot's registration process). E.g.
# Specify a single domain name
MY_DOMAIN_NAME=example.com
# or multiple domains names with this different key (comment out MY_DOMAIN_NAME above if you do):
# MY_DOMAIN_NAMES=example.com anotherexample.net thirdexample.org
[email protected]
-
Now when you want to bring up your server, you will use
docker compose -f docker-compose-letsencrypt.yml up
which will run the server in the foreground so you can make sure everything gets started alright. -
If everything is running, you may want to CTRL+C, run
docker compose -f docker-compose-letsencrypt.yml down
to get to a clean slate, and then rerundocker compose -f docker-compose-letsencrypt.yml up -d
with the added-d
to run the server in the background (in daemon mode) -
Please keep in mind that using the HTTPS method will use the email you specified and the domain name to register the certificate. You can read about the let's encrypt process (using cerbot) over here. The process involves verifying that you are the owner of the domain you have specified and registering you with let's encrypt.
-
THERE IS A RATE LIMIT. So don't keep bringing this server up and down otherwise you will quickly hit a let's encrypt certificate generation limit. To avoid this, for testing purposes you may add
--staging
to the./certbot-auto
command incerbot-nginx/start.sh
which will test whether let's encrypt gives you the certificate.
You may follow these steps if you wish to have a public-facing canarytokens site but would like some basic auth to access it.
git clone https://github.com/thinkst/canarytokens-docker.git
cd canarytokens-docker/nginx
or if you plan on using HTTPS,cd canarytokens-docker/certbot-nginx
sudo htpasswd -c .htpasswd user
whereuser
can be any username you would like to use.sudo chown <user>:<user> .htpasswd
whereuser
is the local linux user- edit the appropriate
nginx.conf
and
server {
...
location ~* (/generate|/manage|/download|/history|/settings|/resources|/legal).* {
auth_basic "Basic Auth Restricted Canrytokens"; <---- ADD
auth_basic_user_file /etc/nginx/.htpasswd; <---- ADD
- edit the appropriate
Dockerfile
and add belowCOPY nginx.conf ...
COPY .htpasswd /etc/nginx/.htpasswd
- rebuild the images using
docker compose build
, restart your docker containers, and enjoy!
Thanks, @mamisano for catching a silly issue using the above 🙏
We are going to track some new features/additions here so that it is quick and easy to see what has been recently added.
-
we now have the capability of sending error logs to a webhook of your choice, hopefully alerting you or your team to the failures as opposed to these errors only living in a log file. Simply supply the corresponding webhook URI in the
ERROR_LOG_WEBHOOK
value in your switchboard.env file. (2021-04-09) -
we've renamed the distributed .env files to
switchboard.env.dist
andfrontend.env.dist
. This ensures that your local configuration doesn't get blown away when you pull changes from the repo. (We still useswitchboard.env
andfrontend.env
for the config, it just means that new clones of the repo require the users to copy/rename the dist files) -
we have added an extra
switchboard.env
calledCANARY_IPINFO_API_KEY
. This allows you to use your ipinfo.io API key if you want to (keep in mind ipinfo.io does have a free tier of up to 1000 requests a day). -
we now have slack support. When you supply a webhook, you simply supply your slack webhook URL. (Thanks to @shortstack).
-
we have added a new environment variable to
frontend.env
calledCANARY_AWSID_URL
which allows you to specify a private or different URL for the AWS ID token. This means you can easily change between accounts. (2018-10-17) -
if you intend to build the image to be run on another system with different architecture, you can build the images with
docker compose build --build-arg ARCH=<target arch>/
, noting the forward slash at the end of the argument. The image will not build correctly if this is not included.