Frequently asked questions

Below are some of the most common questions we receive about the Infection Monkey. If the answer you're looking for isn't here, talk to us on our Slack channel, email us at support@infectionmonkey.com or open an issue on GitHub.

Where can I get the latest version of the Infection Monkey?

For the latest stable release, visit our downloads page. This is the recommended and supported version!

If you want to see what has changed between versions, refer to the releases page on GitHub. For the latest development version, visit the develop version on GitHub.

I updated to a new version of the Infection Monkey and I'm being asked to delete my existing data directory. Why?

The data directory contains the Infection Monkey's database and other internal
data. For the new version of Infection Monkey to work flawlessly, you need to set up a data directory with a compatible structure.

If you would like to save the data gathered from the Monkey's previous runs,
you can make a backup of your existing data directory before deleting it.

How can I use an old data directory?

To use the data stored in a data directory from an older version, reinstall the version of the Monkey Island which matches your data directory's version. Then, copy the backup of your old data directory to the appropriate location.

How long does a single Infection Monkey agent run? Is there a time limit?

The Infection Monkey agent shuts off either when it can't find new victims or it has exceeded the quota of victims as defined in the configuration.

How long does it take to stop all running Infection Monkey agents?

On the Infection Map page, when Kill All Monkeys is clicked, the Agents try to finish execution safely. This can take up to 2 minutes, but will be much shorter on average.

Is Infection Monkey a malware/virus?

Infection Monkey is not malware, but it uses similar techniques to safely
simulate malware on your network.

Because of this, Infection Monkey gets flagged as malware by some antivirus solutions during installation. If this happens, verify the integrity of the downloaded installer first. Then, create a new folder and disable antivirus scan for that folder. Lastly, reinstall the Infection Monkey in the newly created folder.

Reset the Monkey Island password

If you reset the credentials, the database will be cleared. Any findings of the Infection Monkey from previous runs will be lost.
However, you can save the Monkey's existing configuration by logging in with your current credentials and clicking on the Export config button on the configuration page.

On Windows and Linux (AppImage)

When you first access the Monkey Island server, you'll be prompted to create an account.
Creating an account will write your credentials in credentials.json file under data directory. To reset the credentials:

  1. Remove the credentials.json file manually (located in the data directory.

  2. Restart the Monkey Island process:

    • On Linux, simply kill the Monkey Island process and execute the AppImage.
    • On Windows, restart the program.
  3. Go to the Monkey Island's URL and create a new account.

If you are still unable to log into Monkey Island after following the above steps, you can perform a complete factory reset by removing the entire Data directory and then restarting the Monkey Island process.

On Docker

When you first access the Monkey Island server, you'll be prompted to create an account.
To reset the credentials, you'll need to perform a complete factory reset:

  1. Kill the Monkey Island container:
    sudo docker kill monkey-island
    
  2. Kill the MongoDB container:
    sudo docker kill monkey-mongo
    
  3. Remove the MongoDB volume:
    sudo docker volume rm db
    
  4. Restart the MongoDB container:
     sudo docker run \
         --name monkey-mongo \
         --network=host \
         --volume db:/data/db \
         --detach \
         mongo:4.2
    
  5. Restart the Monkey Island container
    sudo docker run \
        --name monkey-island \
        --network=host \
        guardicore/monkey-island:VERSION
    
  6. Go to the Monkey Island's URL and create a new account.

Should I run the Infection Monkey continuously?

Yes, this will allow you to verify that the Infection Monkey identified no new security issues since the last time you ran it.

Does the Infection Monkey require a connection to the internet?

The Infection Monkey does not require internet access to function.

If internet access is available, the Infection Monkey will use the internet for two purposes:

  • To check for updates.
  • To check if machines can reach the internet.

What internet queries does the Infection Monkey perform?

The Monkey performs queries out to the Internet on these separate occasions:

  1. The Infection Monkey agent checks if it has internet access by performing
    requests to pre-configured domains. By default, these domains are monkey.guardicore.com and www.google.com, which can be changed. The request doesn't include any extra information - it's a GET request with no extra parameters. Since the Infection Monkey is 100% open-source, you can
    find the domains in the configuration [here (https://github.com/guardicore/monkey/blob/85c70a3e7125217c45c751d89205e95985b279eb/monkey/infection_monkey/config.py#L152) and the code that performs the internet check here.
    This is not used for statistics collection.
  2. After the Monkey Island starts it sends a GET request with current deployment type to the update server to fetch the latest version and download link for it. This information is used by the Monkey Island to suggest an update if one is available. No information gets collected during this process.
  3. After the Monkey Island starts it sends a GET request to the analytics server with your deployment type and a version number. This information gets collected on the analytics server. It is used to understand which deployment types/versions are no longer used and can be deprecated.

Logging and how to find logs

Downloading logs

Both Monkey Agent and Monkey Island logs can be found in the Infection Map page. Click on the
machine from which you want to download logs and press the "Download log" button on the side panel.
Note that you can only download the Monkey Island log by clicking on the Monkey Island machine in
the Infection Map.

Log locations

If the logs can't be downloaded through the UI for any reason, you can collect the log files
directly from the machine where an Agent or Monkey Island ran.

Monkey Island server logs

The Monkey Island's log file is located in the Data directory.

The log enables you to see which requests were requested from the server and extra logs from the backend logic. The log will contain entries like these:

2022-04-18 13:48:43,914 - pywsgi.py:1226 -      write() - INFO - 192.168.56.1 - - [2022-04-18 13:48:43] "GET /api/agent-binaries/windows HTTP/1.1" 200 21470665 0.293586
2022-04-18 13:48:49,970 - pywsgi.py:1226 -      write() - INFO - 192.168.56.1 - - [2022-04-18 13:48:49] "GET /api/island-mode HTTP/1.1" 200 128 0.003426
2022-04-18 13:48:49,988 - report.py:355 - get_domain_issues() - INFO - Domain issues generated for reporting

It's also possible to change the default log level by editing log_level value in a server configuration file.
log_level can be set to info(default, less verbose) or debug(more verbose).

Infection Monkey Agent logs

The Infection Monkey agent log file can be found in directories specified for temporary files on the machines where it was executed. In most cases, this will be /tmp on Linux and %temp% on Windows. The agent searches a standard list of directories to find an appropriate place to store the log:

  1. The directory is named by the TMPDIR environment variable.
  2. The directory is named by the TEMP environment variable.
  3. The directory is named by the TMP environment variable.
  4. A platform-specific location:
    • On Windows, the directories C:\TEMP, C:\TMP, \TEMP, and \TMP, in that order.
    • On all other platforms, the directories /tmp, /var/tmp, and /usr/tmp, in that order.
  5. As a last resort, the current working directory.

Infection Monkey log file name is constructed to the following pattern: infection-monkey-agent-<TIMESTAMP>-<RANDOM_STRING>.log

The logs contain information about the internals of the Infection Monkey agent's execution. The log will contain entries like these:

2019-07-22 19:16:44,228 [77598:140654230214464:INFO] main.main.116: >>>>>>>>>> Initializing monkey (InfectionMonkey): PID 77598 <<<<<<<<<<
2019-07-22 19:16:44,231 [77598:140654230214464:INFO] monkey.initialize.54: Monkey is initializing...
2019-07-22 19:16:44,231 [77598:140654230214464:DEBUG] system_singleton.try_lock.95: Global singleton mutex '{2384ec59-0df8-4ab9-918c-843740924a28}' acquired
2019-07-22 19:16:44,234 [77598:140654230214464:DEBUG] monkey.initialize.81: Added default server: 10.15.1.96:5000
2019-07-22 19:16:44,234 [77598:140654230214464:INFO] monkey.start.87: Monkey is running...
2019-07-22 19:16:44,234 [77598:140654230214464:DEBUG] control.find_server.65: Trying to wake up with Monkey Island servers list: ['10.15.1.96:5000', '192.0.2.0:5000']
2019-07-22 19:16:44,235 [77598:140654230214464:DEBUG] control.find_server.78: Trying to connect to server: 10.15.1.96:5000
2019-07-22 19:16:44,238 [77598:140654230214464:DEBUG] connectionpool._new_conn.815: Starting new HTTPS connection (1): 10.15.1.96:5000
2019-07-22 19:16:44,249 [77598:140654230214464:DEBUG] connectionpool._make_request.396: https://10.15.1.96:5000 "GET /api?action=is-up HTTP/1.1" 200 15
2019-07-22 19:16:44,253 [77598:140654230214464:DEBUG] connectionpool._new_conn.815: Starting new HTTPS connection (1): updates.infectionmonkey.com:443
2019-07-22 19:16:45,013 [77598:140654230214464:DEBUG] connectionpool._make_request.396: https://updates.infectionmonkey.com:443 "GET / HTTP/1.1" 200 61

Running the Infection Monkey in a production environment

How much of a footprint does the Infection Monkey leave?

The Infection Monkey leaves hardly any trace on the target system. It will leave:

  • Log files in temporary directories.
    • Path on Linux: /tmp/infection-monky-agent-<TIMESTAMP>-<RANDOM_STRING>.log
    • Path on Windows: %temp%\\infection-monky-agent-<TIMESTAMP>-<RANDOM_STRING>.log

What's the Infection Monkey Agent's impact on system resources usage?

The Infection Monkey Agent uses less than a single-digit percent of CPU time and very low RAM usage. For example, on a single-core Windows Server machine, the Infection Monkey Agent consistently uses 0.06% CPU, less than 80MB of RAM, and a small amount of I/O periodically.

If you do experience any performance issues please let us know on our Slack channel or open an issue on GitHub.

What are the system resource requirements for Monkey Island?

Linux

CPU: Intel(R) Xeon(R) CPU @ 2.20GHz or better

CPU Cores: 2

RAM: 4GB

Windows

CPU: Intel(R) Xeon(R) CPU @ 2.20GHz or better

CPU Cores: 4

RAM: 6GB

Is it safe to use real passwords and usernames in the Infection Monkey's configuration?

Absolutely. User credentials are stored encrypted in the Monkey Island server. This information is accessible only to users that have access to the specific Monkey Island.

We advise users to limit access to the Monkey Island server by following our password protection guide.

How do you store sensitive information on Monkey Island?

Sensitive data such as passwords, SSH keys and hashes are stored on the Monkey Island's database in an encrypted fashion. This data is transmitted to the Infection Monkey agents in an encrypted fashion (HTTPS) and is not stored locally on victim machines.

When you reset the Monkey Island configuration, the Monkey Island wipes the information.

How stable are the exploits used by the Infection Monkey? Will the Infection Monkey crash my systems with its exploits?

The Infection Monkey does not use any exploits or attacks that may impact the victim system.

This means we avoid using some powerful (and famous) exploits such as EternalBlue. This exploit was used in WannaCry and NotPetya with huge impact, but, because it may crash a production system, we aren't using it.

After I've set up Monkey Island, how can I execute the Infection Monkey agent?

See our detailed getting started guide.

How can I make the Infection Monkey agent propagate “deeper” into the network?

If you wish to simulate a very “deep” attack into your network, you can increase the propagation depth parameter in the configuration. This parameter tells the Infection Monkey how far to propagate into your network from the “patient zero” machine.

To do this, change the Distance from Island parameter in the configuration's “Basic - Network” tab:

What if the report returns a blank screen?

This is sometimes caused when Monkey Island is installed with an old version of MongoDB. Make sure your MongoDB version is up to date using the mongod --version command on Linux or the mongod -version command on Windows. If your version is older than 4.0.10, this might be the problem. To update your Mongo version:

  • Linux: First, uninstall the current version with sudo apt uninstall mongodb and then install the latest version using the official MongoDB manual.
  • Windows: First, remove the MongoDB binaries from the monkey\monkey_island\bin\mongodb folder. Download and install the latest version of MongoDB using the official MongoDB manual. After installation is complete, copy the files from the C:\Program Files\MongoDB\Server\4.2\bin folder to the monkey\monkey_island\bin\mongodb folder. Try to run the Monkey Island again and everything should work.

Can I limit how the Infection Monkey propagates through my network?

Yes! To limit how the Infection Monkey propagates through your network, you can:

Adjust the scan depth

The scan depth limits the number of hops that the Infection Monkey agent will
spread from patient zero. If you set the scan depth to one, the agent will only
reach a single hop from the initially infected machine. Scan depth does not
limit the number of devices, just the number of hops.

  • Example: In this example, the scan depth is set to two. Host A scans the
    network and finds hosts B, C, D and E. The Infection Monkey agent
    successfully propagates from Host A to Host C. Since the scan depth is 2,
    the agent will pivot from Host C and continue to scan other machines on the
    network. However, if Host C successfully breaches Host E, it will not pivot
    further nor continue to scan or propagate.

Enable or disable scanning the local subnet

You can find the settings that define how the Infection Monkey will scan your network in Configuration -> Network. Each agent will scan its entire local subnet by default, but you can disable this behavior by unchecking the Local network scan button.

Add IPs to the IP allow list

You can specify which hosts you want the Infection Monkey agents to attempt to scan in the Configuration -> Network -> Scan target list section.

Add IPs to the IP block list

If there are any hosts on your network that you would like to prevent the Infection Monkey from scanning or exploiting, you can add them to the list of "Blocked IPs" in Configuration -> Network -> Blocked IPs.

Specify max number of victims to find/exploit

Two settings in Configuration -> Internal -> Monkey allow you to further limit the Infection Monkey's propagation:

  • Max victims to find: This limits the total number of machines that the
    Infection Monkey is allowed to scan.
  • Max victims to exploit: This limits the total number of machines that the
    Infection Monkey is allowed to successfully exploit.

How can I get involved with the project?

Infection Monkey is an open-source project, and we welcome contributions and contributors. Check out the contribution documentation for more information.

About the project 🐵

How did you come up with the Infection Monkey?

Oddly enough, the idea of proactively breaking a network to test its survival wasn't born in the security industry. In 2011, the streaming giant Netflix released Chaos Monkey, a tool designed to randomly disable the company's production servers to verify that they could survive network failures without any customer impact. Netflix's Chaos Monkey became a popular network resilience tool, breaking the network in a variety of failure modes, including connectivity issues, invalid SSL certificates and randomly deleting VMs.

Inspired by this concept, Guardicore Labs developed its own attack simulator - the Infection Monkey - to run non-intrusively within existing production environments. The idea was to test the resiliency of modern data centers against attacks and give security teams the insights they need to make informed decisions and enforce tighter security policies. Since its launch in 2017, Infection Monkey has been used by hundreds of information technology teams from across the world to find weaknesses in their on-premises and cloud-based data centers.