Thursday, July 10, 2025

2025-07-10 - BYU Wi‑Fi captive portal troubleshooting

 

BYU Wi‑Fi Captive Portal Troubleshooting

What happened

I had used BYU Wi‑Fi just fine before, but at some point, the captive portal started acting up.

Every time I opened Edge to trigger the portal, the page would just hang. I started getting warnings like “your connection to this site isn’t secure.” I allowed it anyway, but the page would only half-load—the “Connect” button would start to show, then everything froze. Eventually, it just timed out. No response.

I tried rebooting, updating, reversing settings—nothing changed.


What I tried

  • Checked DNS settings and confirmed they were set to automatic.

  • Cleared the DNS cache and reset Winsock using ipconfig and netsh. Rebooted and tried neverssl.com again—still hung.

  • Verified VPN and proxy were off. I wasn’t running any WireGuard tunnels, and the proxy settings were disabled.

  • Disabled all browser extensions in Edge, used InPrivate mode, and retried the portal. Same result.

  • Tried direct portal URLs, like captive.apple.com. It either said unsupported or “Can’t reach this page.”

  • Forgot the network, reconnected, and tried again—no difference.

  • Double-checked proxy settings. Automatic detection was on, manual proxy was off.

  • Updated the Wi‑Fi adapter driver in Device Manager—still nothing.

At that point, I started preparing for the nuclear option: a full Network Reset and deep stack flush.


Deep dive: what comes next

Before running the reset, I learned something important. Windows needs time to uninstall and reinstall the network stack completely—adapters, services, drivers, all of it. If I reboot too soon, things can break badly. Microsoft recommends waiting about five minutes for it to finish everything under the hood.

Interrupting that process mid-reset can result in missing drivers, failed reinstalls, and total loss of network functionality.


What Network Reset actually does

  • Removes and reinstalls all network adapters (physical and virtual)

  • Wipes out VPN adapters like WireGuard

  • Resets DNS, TCP/IP, proxy settings, and the firewall

  • Gives Windows time to rebuild the entire network stack


What happened to my WireGuard VPN setup

The reset wiped the virtual network adapter and all tunnel configurations. It also removed any routing rules or firewall exceptions that had been set up.

After the reset, I had to:

  • Reopen the WireGuard app

  • Re-import my tunnels

  • Reactivate them

  • Reapply any split-tunnel, kill-switch, or firewall settings I had customized


Captive portal test URL I used

After resetting everything, I used the URL http://captive.apple.com/hotspot-detect.html to trigger the BYU portal. If the network is live and DHCP is functioning, this page should return “Success” or redirect to BYU’s login portal.


What actually fixed it

Fix #1: Full Network Reset

I went to Settings → Network & Internet → Status → Network reset.
Clicked Reset now, then waited a full 5 minutes before doing anything else.
Windows eventually rebooted on its own.

Fix #2: Deep command-line network reset

After reboot, I opened an elevated command prompt and ran:

  • netsh int ip reset

  • netsh advfirewall reset

  • netsh winsock reset

  • ipconfig /release

  • ipconfig /renew

  • ipconfig /flushdns

That flushed everything—old routes, DNS, firewall rules—and gave me a clean slate.

Fix #3: Reconnect and test

  • Reconnected to BYU‑WiFi (forgot and re-added it)

  • Reopened WireGuard, re-imported my tunnel, and activated it

  • Opened Chrome (instead of Edge this time)

  • Went directly to http://captive.apple.com/hotspot-detect.html

This time, I had a valid DHCP lease in the 10.x.x.x range, DNS was automatic, and the BYU login portal finally appeared.


Optional checks I did

  • Verified that I had a real IP address (not 169.254.x.x) using ipconfig /all

  • Temporarily disabled antivirus/firewall to make sure they weren’t blocking HTTP

  • Also tested the portal using http://msftconnecttest.com just in case


Final thoughts

The full reset cleared out everything that could’ve been silently interfering with captive portal authentication—stale DNS entries, old routes, broken VPN adapters, bad proxy settings, all of it.

Once the stack was fully flushed and I used a direct trigger URL, the BYU portal finally loaded.

This has been Truncat3d 00000000111100010100110______________end of line

Wednesday, July 9, 2025

2025-07-09 - Active Directory 4.0 - Joining both Windows 11 Virtual Machines to the Domain

 

How I Solved the Windows 11 Domain Join Fail in My Proxmox AD Lab

I couldn’t get my Windows 11 Enterprise VMs to join the testlab.local domain hosted in Proxmox. I kept running into errors like "Domain not found" or "Active Directory Domain Controller could not be contacted."

Turns out, the issue was DNS not resolving properly over IPv6. Here's how I figured it out and got Win11Ent‑1 to finally join the domain.


Background

  • Lab is running on a Lenovo M910 with Proxmox

  • Domain: testlab.local

  • Domain Controller: Server 2025 VM WIN‑BCCTIROV6RS
    → Static IP: 192.168.50.191

  • All client VMs are pointed to that IP for DNS

  • FSMO roles confirmed using netdom query fsmo

  • Two Windows 11 Enterprise VMs (only one runs at a time with the DC)


Symptoms

  • nslookup testlab.local on Win11Ent‑1 returned a non-existent domain and showed an IPv6 address

  • Attempting to join the domain failed with “Domain not found” errors

  • DNS was definitely running: the zone existed and all SRV records were in place



What I Did

How I Tried to Add the Workstation to the Domain

Before attempting to join the domain, I made sure the workstation was pointing to the correct DNS server:

  • Set DNS Server to the Domain Controller IP

  • Press Win + R, type ncpa.cpl, press Enter

  • Right-click the Ethernet adapter, select Properties

  • Select Internet Protocol Version 4 (TCP/IPv4), click Properties

  • Select Use the following DNS server addresses:

  • Preferred DNS server: 192.168.50.191

  • Leave alternate blank

  • Click OK, then close all windows

Attempting to Join Windows 11 Enterprise to the Domain

  • Press Win + R, type sysdm.cpl, press Enter

  • In System Properties, go to the Computer Name tab, click Change…

  • Select Domain > entered: testlab.local

  • Got error: “An Active Directory Domain Controller (AD DC) for the domain ‘testlab.local’ could not be contacted.”

    Ensure that the domain name is typed correctly. If the name is correct, click Details for troubleshooting information.

  • nslookup testlab.local indicated that Windows was querying an IPv6 DNS server (instead of my IPv4 DNS at 192.168.50.191)

  1. Fixing the problem, Verified DNS server is reachable

Win11Ent-1 > Open Command Prompt > ping 192.168.50.191

  • 100% success

  1. Checked DNS port connectivity

Win11Ent-1 > PowerShell > Test-NetConnection 192.168.50.191 -Port 53

  • Port was open

  1. Forced the VM to use only IPv4 DNS

Win11Ent-1 > Win + R > ncpa.cpl
Right-click Ethernet adapter > Properties
Uncheck Internet Protocol Version 6 (TCP/IPv6)

Click on Internet Protocol Version 4 (TCP/IPv4) > Properties
Set DNS server manually: 192.168.50.191
Clear alternate DNS field

  1. Flushed and re-registered DNS

Win11Ent-1 > Run Command Prompt as Admin:

ipconfig /flushdns
ipconfig /registerdns
  1. Tested DNS directly against the server

Command Prompt > nslookup testlab.local and also nslookup testlab.local

192.168.50.191

  • Returned 192.168.50.191 as expected

  • Still said "Server: Unknown" but that’s fine for now

  • Entered > echo %userdomain%

  • Returned TESTLAB (finally!)

  1. Tried joining the domain again

Win + R > sysdm.cpl
Computer Name tab > Change…
Select Domain and enter: testlab.local
It worked this time and prompted for credentials





✅ Success
Disabling IPv6 fixed it. Domain join prompt came up.

I used:

  • Username: Administrator

  • Password: WinServ2025

Got the “Welcome to the testlab.local domain” message. 

Restart Required After Domain Join

After successfully joining a Windows 11 workstation to the domain, a restart is required before domain credentials can be used to log in.

Here’s what I did:

  • Joined the domain:
    Win + R > sysdm.cpl > Computer Name tab > Change… > Select Domain > enter testlab.local

    Enter credentials: TESTLAB\Administrator > Password: WinServ2025

  • Got the success message saying the machine was added to the domain.

  • Prompt appeared:

    “You must restart your computer to apply these changes.”

So I followed the prompt:

  • Restarted the workstation.

  • Upon Rebooting, I clicked Other User and entered TESTLAB\Administrator as the username and WinServ2025 as the password, which successfully logged me into the domain account. 

  • It took a minute but it logged into the account.

  •  I set up the other machine and it was very straightforward after that. Changed the DNS address statically, turned off IPv6, entered the nslookup TESTLAB and echo %userdomain% commands and added the machine to the domain, and got the right results!


Summary

  • DNS was fine on the server.

  • Client prioritized IPv6 and tried resolving via ::1 and a public IPv6 resolver.

  • That broke DNS resolution until I disabled IPv6.

  • Fix was: disable IPv6, force IPv4 DNS, flush and register, reattempt domain join.

If your lab VM won’t join a domain but DNS looks fine, check if IPv6 is getting in the way.

This has been Truncat3d 00000000111100010100110______________end of line

2025-05-28 - Active Directory 3.0 - Setting Up Virtual Machines to Join the Active Directory Domain

Setting Up Windows 11 Enterprise LTSC VMs and Windows Server 2025 Domain Controller in Proxmox

        I went to https://www.microsoft.com/en-us/evalcenter/evaluate-windows-11-enterprise and downloaded the LTSC version of Windows 11 Enterprise since what I am using these ISOs for is very short-lived and not wide in scope. I will be doing this on a server that I don’t believe has much power, even though my friend already in IT said it does. Now to create the VMs. I think maybe I’ll set up two for now. I will probably create a few more but not run them at the same time.

        Before I went any further I wanted to find out how much time I had left on my evaluation of Windows Server. I have been running it for several months now, not even sure how many but I just assumed I was coming up on the end soon. Wouldn't have surprised me if I had a week left because it seemed like it was such a long time ago. But I looked up the command to find out how many days or minutes were remaining on my evaluation in windows server, I don't know if this works in other OS's but I entered:

    slmgr /dlv

          This command gave me the info:

      I was told I had 181,899 minutes and 127 days left on the Windows Server evaluation. It also showed I had a remaining Windows rearm available. After looking into that, I found I can refresh the evaluation one time before needing to reinstall. That sounded like something I wanted to keep in mind.

        I hadn’t uploaded the OS ISOs to Proxmox yet, so I went to the Proxmox web interface:
Datacenter > Proxmox node > Local (proxmox) storage (where OS images go) > ISO Templates > Upload > Select File > navigated to my file server to pick the ISOs.
It took quite a while, barely passing 6% after five minutes.

        I first tried uploading the Windows 11 ISO via the Proxmox web interface, but it kept failing with error 0. Switching to SCP from my Windows laptop terminal to copy the ISO directly to the Proxmox server also failed, despite several tries.

        I discovered my 256GB NVMe drive was partitioned by Proxmox with only about 68GB for root partition, which was almost full. Running df -h and du on Proxmox showed most root space was taken up by VM backups in /var/lib/vz/dump. Apparently, my recent work to protect the Windows Server evaluation from expiring caused this space issue.

        I deleted older backups in /var/lib/vz/dump, keeping only the latest one, and adjusted Proxmox backup retention to keep just one backup going forward. This freed up significant space, and now I had plenty of room for uploading ISOs and creating VMs.

To check disk usage in /var/lib/vz and find largest folders:
du -h --max-depth=1 /var/lib/vz | sort -hr

To list backups sorted newest first:
ls -t /var/lib/vz/dump/

To delete all backups except the newest one:
ls -t /var/lib/vz/dump/vzdump-qemu-100-*.vma.zst | tail -n +2 | xargs -r rm -f

        This command deletes all .vma.zst backup files except the latest.

        I also learned how Proxmox splits storage between root and volume groups, and how VM disks, backups, ISOs, and system files are stored separately. This helped me understand why I thought I had enough space but actually ran out on root.

        I retried uploading via the web interface, and this time it finished within a few minutes. Now I’m thinking about shrinking the thin provision size for my Windows Server VM since it’s currently 50GB and might not need that much. I thought I had 256GB total, but partitioning changes the available space significantly.


2025-06-10

        Yesterday I spent hours trying to get the first Windows 11 VM to boot properly in Proxmox. I must’ve recreated the VM at least 15 times chasing the same errors. Every time I tried UEFI, I got the “BdsDe: failed to load Boot0002 UEFI EMU QEMU HARDDISK” error or similar boot failures. When I switched to SeaBIOS, I could reach the installer but setup complained about not meeting Windows 11 system requirements.

        At first, I suspected specs—cores, RAM, or disk size—so I tested different virtual disk sizes: 30GB, 60GB, 64GB. I needed to keep drives small to fit multiple workstation VMs on my limited storage. The errors kept happening, and I started wondering if I really needed more than one VM.

        I imagined a job interview: “Why do you want this job if you can do advanced work like this?” they’d ask. Then, “You mean you only added one VM?” Me, smiling, “Yep!” “Well, we’ll let you know… don’t call us, we’ll call you!”

        I also experimented with many VM creation settings: toggling QEMU Agent, switching between SATA, IDE, SCSI, and VirtIO Block for system > Bus/Device, changing RAM. Nothing fixed it. I followed many instructions—some swore by VirtIO Block, ChatGPT by IDE, others recommended SCSI. I even mistakenly chose SATA once. None of these solved the two persistent errors blocking Windows install.

        Each time I recreated the VM, I followed the same process:
Datacenter > my node > right-click > Create VM.
In the wizard, selected OVMF (UEFI) under System > BIOS and added TPM 2.0 support.
For storage controller, chose VirtIO SCSI single.
Mounted Windows 11 ISO and VirtIO drivers ISO under CD/DVD.
Hard drive bus set to SCSI with 60GB or 64GB.

        Eventually, I noticed the “Press any key to boot from CD or DVD…” prompt flashed for a fraction of a second before the UEFI error. I kept missing it and letting it timeout. Once I caught it and pressed a key fast enough, the Windows 11 installer launched successfully. All previous errors were just because I wasn’t actually booting the installer.

        After wasting so much time testing drive sizes, tweaking settings, switching BIOS types, and trying to outsmart Proxmox, it was that simple. So if you get “BdsDe” boot failures or your Windows VM won’t boot from ISO, make sure you’re not ignoring the “Press any key” prompt like I did.


        Trevor, my second instructor at MTECH, helped me with this. When the installer finally worked, I thought, I need two VMs, and if I can’t recreate this, I’m stuck. I deleted the VM and started fresh to reproduce errors for Trevor.

        He kept suggesting starting fresh, and once I stepped away, he started the VM. Suddenly it worked. We reviewed every step from another guide online, sure we had solved the issue. We agreed to document every detail. Trevor suggested deleting that VM too since it worked now, and I reluctantly agreed.

We recreated the VM but it still didn’t work for about another half hour.

From there  went BROWSE >

I selected the drive that had VIRTIO-WIN-0.1.271 loaded in >

I selected VIOSTOR >

Selected W11 > 

AMD64 > pressed OK >

Selected RED HAT  VIRTIO  SCSI CONTROLLER > pressed INSTALL >

 selected the available disk in the case of whichever VM I was making (this is a different one)  > pressed NEXT >

From there, the installation proceeded normally. I thought it was weird the drive didn’t show up immediately like before, but it worked. Since drive space was tight, I decided to recreate the first VM rather than jump straight to the second one. For safety, I kept the first VM Trevor and I made successfully—no way I wanted to lose that progress and be stuck again. After all, I needed to keep moving forward.

        However, when I tried making the second VM, I ran into the same issue: the drive wouldn’t show up at the window prompting me to select the install location. I repeated the process of loading the VirtIO driver but was confused—why hadn’t I needed to do this for the previous VMs?

        After many failed attempts and Trevor also stumped, I researched online and found three differences in my setups:

  • The successful VM didn’t use QEMU Agent

  • It used VirtIO Block for the hard drive bus

  • I had set disk sizes at 64GB, then 60GB (which worked), because both Windows 11 VMs had to fit within Proxmox’s 125GB partition limit (about 12GB actually allocated per VM)

        On my next attempt, I aimed for two VMs with only 30GB each, but after failing many times and retrying 60GB again, I stuck with 60GB.

        Online resources clarified the issue: VirtIO Block requires the VIOSTOR driver to show the drive during installation, while SCSI requires VIOSCSI. I had been using SCSI without understanding this.

        VirtIO Block is simpler and faster but needs VIOSTOR; SCSI is more flexible with advanced features and uses VIOSCSI.

        When I recreated the VM following this knowledge, it worked. I don’t remember if I used VirtIO Block or SCSI in the end, but I suspect VirtIO Block with VIOSTOR since I knew where to find those drivers and didn’t see VIOSCSI drivers afterward.

        Windows installed! I’m glad I kept that initial 60GB VM. Now I feel confident I know what’s needed to get Windows 11 installed on Proxmox VMs. Both VMs have 60GB disks, fitting within my 125GB limit. I don’t see a need for more VMs right now, so I’m sticking with these two and moving on.

        Setting Up Windows 11 Enterprise VMs for Active Directory Domain Join on Windows Server 2025 (Proxmox Lab)

I’m building an Active Directory lab using Proxmox with:

  • A Windows Server 2025 VM promoted to domain controller with AD DS and DNS installed

  • Two Windows 11 Enterprise LTSC VMs ready to join the domain

What I’ve done so far:

  • Confirmed the domain controller VM is running and promoted to domain TESTLAB

  • Checked the domain controller’s IP address inside the server with ipconfig — it’s 192.168.50.191

  • Powered on the two Windows 11 VMs, named Win11Ent-1 and Win11Ent-2

  • Noted their IPs: 192.168.50.51 and 192.168.50.106 respectively

  • Changed the DNS server on both Windows 11 VMs to point to the domain controller’s IP 192.168.50.191 by:

    Pressing Win + R, typing ncpa.cpl, pressing Enter
    Right-clicking Ethernet adapter > Properties
    Selecting Internet Protocol Version 4 (TCP/IPv4) > Properties
    Setting Preferred DNS server to 192.168.50.191 > OK > Close

  • Verified network connectivity between the clients and server (e.g., ping, nslookup)

  • Preparing to join the Windows 11 VMs to the TESTLAB domain next.

This has been Truncat3d 00000000111100010100110______________end of line

Thursday, July 3, 2025

2025-07-03 - How to Stop “Open File - Security Warning” When Opening Network Shortcuts

How to Stop “Open File - Security Warning” When Opening Network Shortcuts

If you’ve recently started seeing the annoying “Open File - Security Warning” every time you open a movie shortcut stored on a file server, you’re not alone. This is a common problem caused by Windows treating files from network locations as untrusted, especially after recent security updates.

Here’s how I tackled the issue step-by-step, and what worked.


Why This Happens

  • Windows flags files from network shares as potentially unsafe.

  • This triggers a security warning every time you open shortcuts or files.

  • It can appear suddenly after Windows updates tighten security.

  • Running shortcuts from a file server or mapped drive triggers this.


What I Tried and Why

Unblocking the Shortcut File

  • Right-click shortcut > Properties > Unblock checkbox

  • This often works for downloaded files but doesn’t help much for network shares.

Moving the Shortcut Locally

  • Copying the shortcut to the local PC removes the warning.

  • Not practical if you want to keep files centralized on the server.

Adding Server to Trusted Zone via Internet Options

  • Tried going to inetcpl.cpl > Security > Local Intranet > Sites > Advanced

  • No “Advanced” or “Sites” button visible — likely due to Group Policy restrictions.


The Solution: Adding the File Server IP to Local Intranet via Registry

Since GUI options were locked down, I edited the registry directly:

  1. Open Registry Editor (regedit).

  2. Navigate to:

    pgsql

    HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Internet Settings\ZoneMap\Ranges
  3. Create a new key called Range1.

  4. Inside Range1, add a new String Value named :Range and set it to your server’s IP, for example:

    192.168.50.153
  5. Add a new DWORD (32-bit) Value named file and set it to 1 (leave as Hexadecimal).

  6. Close Registry Editor and reboot.


How to Find Your Server Name or IP for This

  • Check your shortcut properties for a UNC path (\\ServerName\Share), or

  • If you use a mapped drive (like S:), run net use in Command Prompt to see the real UNC path.

  • Use the server name or IP from that UNC path in the registry keys.


Result

After reboot, the security warning disappeared when opening shortcuts on the network drive.


Summary

  • The warning is Windows protecting you from untrusted network locations.

  • GUI options to add trusted sites may be blocked by domain policies.

  • Editing the registry is an effective workaround.

  • Always back up your registry before making changes.


Feel free to share this if you’re frustrated by this recent Windows annoyance. I was too until I cracked it.

This has been Truncat3d 00000000111100010100110______________end of line

2025-07-10 - BYU Wi‑Fi captive portal troubleshooting

  BYU Wi‑Fi Captive Portal Troubleshooting What happened I had used BYU Wi‑Fi just fine before, but at some point, the captive portal star...