Passwordless RDP Authentication for On-Prem Servers with Smart Cards (FIDO2 Security Key)
Hello Everyone, in my previous blog, I discussed how to use FIDO2 Security Key Passwordless Authentication with Entra or Hybrid Joined devices for Remote Desktop Connection. In this blog, we will discuss how a FIDO2 Security Key can be used as a smart card for Remote Desktop Connection Protocol on on-prem Active Directory domain-joined servers. If you have not gone through my last article, you can click here. Many people have written several articles and blogs on similar topics, so I apologize if you find it duplicate. Recently, a customer asked me whether it was possible to authenticate using a FIDO2 security key to connect to a remote on-prem domain joined server via RDP. I successfully enabled it for them and decided to write a blog on the topic. This is a great use case for administrators who frequently use high-privileged groups like Domain Admins and Enterprise Admins, as they should adopt passwordless authentication for secure day-to-day server management using smart card-based authentication. Basically, what I will be demonstrating here is Kerberos PKINIT (Public Key Cryptography for Initial Authentication in Kerberos), which is an extension of the Kerberos Authentication protocol that allows users to authenticate using certificates. There are some vendors who facilitate PIV (Personal Identification Verification) in security keys for strong authentication. With PIV, we can use the security key to store certificates for authentication. So, let’s discuss how we can use this feature to go passwordless by accessing the remote desktop of servers. First of all, we need a PKI infrastructure or a Certificate Server, which we can deploy on Windows Server by adding the Certificate Services role or using an existing one. Once we have a certificate server, we can configure the certificate templates for users and domain controllers. In order to use RDP for servers, both the client and domain controller (DC) need to have a valid certificate as they use mutual authentication. We can issue a “User or Smart Card Logon” certificate to users and a “Kerberos Authentication” certificate to DCs. Before we get into settings and configuration, let’s understand how Kerberos PKINIT works at a high level, which will further help us understand the entire process of this activity. Kerberos Public Key Cryptography for Initial Authentication (PKINIT) in the Kerberos protocol enables the use of public key cryptography in the initial authentication exchange. It uses X.509 certificates (Smart Card) in place of a password to authenticate against the authentication server. The key components involved here are: The Domain Client running Windows 10/11, who wants to access a remote server over RDP. The Domain Controller (KDC) running on at least Windows Server 2016, which authenticates users in Active Directory. The Domain Member Server running on at least Windows Server 2016, which is the target system the client wants to connect to. A Certificate Authority (CA) running on at least Windows Server 2016, which issues Kerberos Authentication and User or Smart Card Logon certificates. Authentication Flow: The Client selects a smart card during authentication, which sends an AS-REQ (Authentication Service Request) to the Domain Controller (KDC) containing the user’s X.509 certificate. It essentially signs the current time with its private key. The Domain Controller validates the request by checking the times using the user’s public key. Once the Domain Controller completes the validation, it issues a TGT (Ticket Granting Ticket) signed by the KDC certificate using its private key as an AS-REP (Authentication Service Reply) response. The Client validates the TGT by verifying the KDC’s certificate. Once the Client has the TGT, it can proceed to request a service ticket to connect to the target server. It is important to note that if we have an Enterprise or AD-integrated Certificate Authority, the Root CA or Issuing CA's certificate will be automatically added to the Trusted Root Certification Authorities store in domain-joined systems. In case we use a standalone CA, we must manually add its certificate to the client machine’s Trusted Root Certification Authorities store. Now, let’s go through the whole process step by step. We will first start by creating a template for the Domain Controller (DC) certificate and later for User certificates. We will also see how to configure Group Policy for certificate auto-enrollment to issue certificates to users and DCs. Issue Kerberos Authentication Certificate to Domain Controller: Go to the Certificate Server and open the Certificate Authority console. Click on Templates and then click on Manage. The Domain Controller (DC) requires the KDC Authentication certificate (1.3.6.1.5.2.3.5) EKU and Server Authentication (1.3.6.1.5.5.7.3.1) EKU. Select the Kerberos Authentication template. Select the Kerberos Authentication template and create a duplicate template. Next, do not make any changes to the certificate template except for assigning a name under the General tab. Go to the Security tab and ensure that only the Domain Controllers group is added with Read, Enroll, and Autoenroll permissions selected. Under the Subject Name tab, ensure the DNS checkbox is selected. Next, we need to issue the Kerberos Authentication certificate template to make it available for Domain Controllers (DCs) to request certificates. Group Policy Configuration: Next, we need to create a Group Policy for certificate auto-enrollment and link it to the Domain Controllers OU. Open the GPO and go to Computer Configuration\Windows Settings\Security Settings\Public Key Policies. Edit Certificate Service Client – Auto Enrollment Properties and select Renew Expired Certificate and update Certificate options as shown in picture below. Go to Computer Configuration\Windows Settings\Security Settings\Public Key Policies and set "Certificate Services Client – Certificate Enrollment Policy" to Enabled. We also need to create another Group Policy Object (GPO) and link it to the domain to enable additional policies for the Client & Server to accept smart card authentication for RDP connections. This policy will determine how the system should behave when the smart card is removed. In this case, it will be set to disconnect the session. Go to Computer Configuration\Windows Settings\Security Settings\Security Options and enable "Define this policy setting" and select option "Disconnect if a Remote Desktop Services Session" Next, we will enable the use of smart cards by setting the policy "Allow ECC certificates to be used for logon and authentication" under: Computer Configuration\Administrative Templates\Windows Components\Smart Card Create Smart Card Logon Certificate Template for Client: Now, go back to the Certificate Server, open the Certificate Authority console, and open the Manage console by right-clicking on Certificate Templates. Select the Smart Card Logon template, right-click, and choose Duplicate Template. Under the Compatibility tab, set: Certificate Authority to Windows Server 2016 Certificate Recipients to Windows 10/Windows Server 2016 Next, go to the General tab and assign a name of your choice. This is the same certificate that the user will see when they issue a smart card certificate to be stored in the FIDO2 Security Key. Go to the Request Handling tab and: Under Purpose, select "Include symmetric algorithms allowed by the subject." Enable "For automatic renewal of smart card certificates, use existing key if a new key cannot be created." To ensure the certificate is saved in the FIDO2 Security Key during the request, select "Prompt the user during enrollment and require user input when the private key is used." Note: I tested the "Prompt the user during enrollment" option, but it did not work. Next, go to the Cryptography tab and: Under Provider Category, select "Key Storage Provider." In Algorithm Name, choose "ECDH_P384" (assuming you meant P384, as P383 is not a standard option). Under Cryptographic Provider, select "Request must be one of the following providers" and choose "Microsoft Smart Card Key Storage Provider." Change Request Hash to "SHA256." Next, go to the Security tab and: Ensure the group containing Admins is added with Read and Enroll permissions. Optionally, enable Autoenroll if needed. Finally Click on OK to save the new template. Let the Group Policy refresh automatically or manually force it by running gpupdate /force. Once refreshed, the Domain Controller (DC) should receive a new Kerberos Authentication Certificate. Enroll Client Smart Card Certificate: On a Windows 10/11 device, Open Command Prompt and run “certreq -enroll "<SmartCard Certificate Template Name>" Ensure the appropriate Smart Card certificate is selected, then click Next to proceed with the certificate issuance. Insert the FIDO2 Security Key. Once the system detects the security key, it will prompt you to enter the PIN to store the certificate The process of storing certificate in FIDO2 Security key completes and now we can test accessing server using RDP with security key Testing: Open mstsc.exe and enter the target server’s FQDN. If the security key is detected by the system, it will prompt you to use the smart card for login. Enter the PIN, and it should allow you to sign in successfully Once authentication is successfully completed, RDP should load the desktop. Troubleshooting: During my research and extensive testing in my lab, I encountered few errors when enrolling certificates on the FIDO2 Security Key PIV. One of the issues I faced was the smart card showing as locked. After troubleshooting, I found that sometimes the FIDO2 security key device driver provided by the vendor is not installed properly. Ensure that you follow the security key provider’s installation guide and install the latest driver on both the client and target server. Another common error I encountered was: "The requested key container does not exist in smart card." This issue typically occurs when the FIDO2 Security Key driver is not installed properly on the target server. Again, refer to the FIDO2 security vendor’s documentation to install the correct driver. . Note: Ensure that when you attempt RDP to the target server, the user for whom you issued the smart card certificate is added to the "Remote Desktop Users" group on the target server. If you are looking to use RDP from the internet, we have the option of KDC Proxy to use. You can refer this article here for more details. I hope you found this blog useful in achieving passwordless authentication even for on-prem Active Directory domain-joined critical servers and I would like to thank you for reading this blog. Hopefully I will be back soon with some more interesting blogs.Feature Request, Let us use one set of User/Pass for multiple Related domains in Edge password mgr
This is a useful feature that LastPass has. this would prevent from creating multiple duplicate entries in Edge password manager and also manage logins more efficiently. in LastPass it's called Equivalent domains, the description says: "Add domains that use the same login service. We've already listed popular sites that use shared credentials across domains under their control. For example: amazon.com and its local variations, or gmail.com and other google products." it's pretty clear, instead of saving 2 password entries for Facebook.com and Messenger.com , we use the same set of username and password for both websites. this also makes it easy when user changes password for one website, because the other one will be updated automatically. here are more real and important examples: a bit hard to share it through Edge feedback and also keep the proper formatting but i'll try. Deleted[Solved] How do I properly add a wildcard domain to Edge tracking prevention exception list?
Found the solution, which was there in front of my eyes all this time but I didn't pay enough attention. Edge shows the correct syntax [*.]example.com so entering [*.]live.com is acceptable and fixes the issue! lesson learned: always pay attention to the info on the screen.
Deploy Open Web UI on Azure VM via Docker: A Step-by-Step Guide with Custom Domain Setup.
Introductions Open Web UI (often referred to as "Ollama Web UI" in the context of LLM frameworks like Ollama) is an open-source, self-hostable interface designed to simplify interactions with large language models (LLMs) such as GPT-4, Llama 3, Mistral, and others. It provides a user-friendly, browser-based environment for deploying, managing, and experimenting with AI models, making advanced language model capabilities accessible to developers, researchers, and enthusiasts without requiring deep technical expertise. This article will delve into the step-by-step configurations on hosting OpenWeb UI on Azure. Requirements: Azure Portal Account - For students you can claim $USD100 Azure Cloud credits from this URL. Azure Virtual Machine - with a Linux of any distributions installed. Domain Name and Domain Host Caddy Open WebUI Image Step One: Deploy a Linux – Ubuntu VM from Azure Portal Search and Click on “Virtual Machine” on the Azure portal search bar and create a new VM by clicking on the “+ Create” button > “Azure Virtual Machine”. Fill out the form and select any Linux Distribution image – In this demo, we will deploy Open WebUI on Ubuntu Pro 24.04. Click “Review + Create” > “Create” to create the Virtual Machine. Tips: If you plan to locally download and host open source AI models via Open on your VM, you could save time by increasing the size of the OS disk / attach a large disk to the VM. You may also need a higher performance VM specification since large resources are needed to run the Large Language Model (LLM) locally. Once the VM has been successfully created, click on the “Go to resource” button. You will be redirected to the VM’s overview page. Jot down the public IP Address and access the VM using the ssh credentials you have setup just now. Step Two: Deploy the Open WebUI on the VM via Docker Once you are logged into the VM via SSH, run the Docker Command below: docker run -d --name open-webui --network=host --add-host=host.docker.internal:host-gateway -e PORT=8080 -v open-webui:/app/backend/data --restart always ghcr.io/open-webui/open-webui:dev This Docker command will download the Open WebUI Image into the VM and will listen for Open Web UI traffic on port 8080. Wait for a few minutes and the Web UI should be up and running. If you had setup an inbound Network Security Group on Azure to allow port 8080 on your VM from the public Internet, you can access them by typing into the browser: [PUBLIC_IP_ADDRESS]:8080 Step Three: Setup custom domain using Caddy Now, we can setup a reverse proxy to map a custom domain to [PUBLIC_IP_ADDRESS]:8080 using Caddy. The reason why Caddy is useful here is because they provide automated HTTPS solutions – you don’t have to worry about expiring SSL certificate anymore, and it’s free! You must download all Caddy’s dependencies and set up the requirements to install it using this command: sudo apt install -y debian-keyring debian-archive-keyring apt-transport-https curl -1sLf 'https://dl.cloudsmith.io/public/caddy/stable/gpg.key' | sudo gpg --dearmor -o /usr/share/keyrings/caddy-stable-archive-keyring.gpg curl -1sLf 'https://dl.cloudsmith.io/public/caddy/stable/debian.deb.txt' | sudo tee /etc/apt/sources.list.d/caddy-stable.list sudo apt update && sudo apt install caddy Once Caddy is installed, edit Caddy’s configuration file at: /etc/caddy/Caddyfile , delete everything else in the file and add the following lines: yourdomainname.com { reverse_proxy localhost:8080 } Restart Caddy using this command: sudo systemctl restart caddy Next, create an A record on your DNS Host and point them to the public IP of the server. Step Four: Update the Network Security Group (NSG) To allow public access into the VM via HTTPS, you need to ensure the NSG/Firewall of the VM allow for port 80 and 443. Let’s add these rules into Azure by heading to the VM resources page you created for Open WebUI. Under the “Networking” Section > “Network Settings” > “+ Create port rule” > “Inbound port rule” On the “Destination port ranges” field, type in 443 and Click “Add”. Repeat these steps with port 80. Additionally, to enhance security, you should avoid external users from directly interacting with Open Web UI’s port - port 8080. You should add an inbound deny rule to that port. With that, you should be able to access the Open Web UI from the domain name you setup earlier. Conclusion And just like that, you’ve turned a blank Azure VM into a sleek, secure home for your Open Web UI, no magic required! By combining Docker’s simplicity with Caddy’s “set it and forget it” HTTPS magic, you’ve not only made your app accessible via a custom domain but also locked down security by closing off risky ports and keeping traffic encrypted. Azure’s cloud muscle handles the heavy lifting, while you get to enjoy the perks of a pro setup without the headache. If you are interested in using AI models deployed on Azure AI Foundry on OpenWeb UI via API, kindly read my other article: Step-by-step: Integrate Ollama Web UI to use Azure Open AI API with LiteLLM Proxy
Windows Server Essentials 2019 shutting down on Domain Environment
Hello, the company that I work for uses Windows Server Essentials 2019 in our products, but we have been receiving complaints from our customers that servers are shutting down weekly when they have this machines added to the Domain environment. We can see that service SILSVC is the responsible for this action, but we don't know the reason. Is there a limitation of Essentials with Domain, or maybe a misconfiguration either on the Windows Roles or Domain configuration?SS38: screenshot based warning - color tabs - reload - download status on tabs
Suggestion (SS): 38 Classification: warning - download on tabs PRIORITY IN MY OPINION: 4 on a scale from 1 (low) to 10 (high) Edge devs, i wrote you before via feedback app, but i continue the topic here, so that other chromium based devs can read it too. We talked about reload by domain and why such feature is cool and useful. now a similar suggestion is that you detect all website that has the same identical layout (so screenshot based). if i click via menu bar or right click menu "warn me based on this website" you color all tabs that have the similar layout/design. related to the picture, you warn me if the link has not be saved correctly on all other tabs. now about this website we generally** can see such problems via the icon in the tab, but there are a lot other examples, where we need to check manually, otherwise we batch close not completed websites. **there are still situations where the icon is not a valid method to know if the website (in this case webarchive) completed the task correctly or not. this was another example we talked via feedback app where maybe a better integration between "smart screenshot warning" and "reload tab" should happen. another suggestion related to this is to show me on the tab an icon --> that marks you are downloading files from such tab and once the download is done, you color it in green.
