STATUS_NDIS_OFFLOAD_CONNECTION_REJECTED Fix (0XC0231012)

Quick answer

Run netsh int tcp set global chimney=disabled in an admin Command Prompt, then reboot. That kills the offload feature causing the error.

What's going on here

This error pops up on Windows machines—usually Windows 10 or Windows Server 2016/2019—when the TCP Chimney Offload feature tries to hand off a connection to your network card, but the NIC or its driver rejects it. Microsoft introduced Chimney Offload years ago to reduce CPU usage by moving TCP processing to hardware. Sounds good on paper, but in practice, many NICs—especially older Realtek, Broadcom, or even some Intel adapters—simply can't handle it reliably. You'll see this when a heavy download stalls, an RDP session drops, or an application that streams data over TCP (like a database sync or backup job) hangs mid-transfer.

I've seen this on a Dell PowerEdge server running Windows Server 2016 with an integrated Broadcom NetXtreme card. The error logged in Event Viewer under System with source 'NetBT' or 'Tcpip' and ID 4226 or 4231. Don't waste time chasing phantom network issues—this is a driver/hardware offload incompatibility, pure and simple.

Fix it in 3 steps

Disable TCP Chimney Offload
Open Command Prompt as Administrator. Run:
```
netsh int tcp set global chimney=disabled
```
This turns off the entire offload engine. No reboot needed immediately, but do reboot after you verify other settings.
Confirm the change
Run:
```
netsh int tcp show global
```
Look for 'Chimney Offload State' — it should say 'disabled'. If it still says 'enabled', your group policy or antivirus (looking at you, Symantec Endpoint Protection) might be overriding the setting. Check local group policy under Computer Configuration > Administrative Templates > Network > TCPIP Settings > 'Enable TCP Chimney Offload' and set it to Disabled.
Disable RSS and VMQ too (if Chimney alone didn't fix it)
Some netcards also trip over Receive Side Scaling (RSS) or Virtual Machine Queue (VMQ). Run:
```
netsh int tcp set global rss=disabled
```
Then in Device Manager, right-click your NIC, go to Advanced tab, and find 'Virtual Machine Queues' (if present). Set it to Disabled. Reboot.

Alternative fixes if you can't disable Chimney

Maybe your corporate policy requires Chimney enabled for some legacy app. Rare, but possible. In that case, update the NIC driver. Go to your motherboard or server vendor's site—don't trust Windows Update for this. For Broadcom adapters on Dell hardware, grab the driver pack from Dell's support page, version 17.4.1 or newer. For Realtek, use their official driver 10.045 or higher from the Realtek site.

If driver update fails, swap the network card. I've had good luck with Intel PRO/1000 or i350 cards—they handle offload like champs. Or, if you're on a virtual machine (Hyper-V or VMware), disable VMQ on the virtual switch. In Hyper-V Manager, open Virtual Switch Manager, select your external switch, and uncheck 'Enable virtual machine queue'. That stops the error dead.

Prevention tip

Once you've disabled Chimney Offload, this error won't come back unless you re-enable it. But to avoid similar offload headaches, turn off all TCP offload features proactively on any older hardware. On new builds, test with Chimney enabled first—if you see Event ID 4226 within a week, disable it. For servers, I always disable Chimney and RSS during initial provisioning. Saves hours of troubleshooting later.

One more thing: Some third-party firewalls (like Comodo or ZoneAlarm) interfere with offload settings. If the error persists after disabling Chimney, temporarily uninstall the firewall to test. Add it back after confirming the network works.

You'll be back to full speed without those cryptic connection drops. If you're still hitting this on Windows 11 or Server 2022, Chimney Offload is deprecated and usually off by default—check for third-party software injecting into the TCP stack. Run a clean boot to isolate it.