While trying to troubleshoot some issues with a user’s flaky VPN connection (they could get connected but couldn’t ping anything on the company network), my first guesses were that it was either a routing issue and that IPv6 might be the culprit or that it could be an MTU issue. In searching for the answers on how to disable IPv6, I had stumbled across a blog article that, while seemingly helpful at the time, had some misleading information on it that caused some rather undesirable results. It gave me an incorrect registry value setting of “0xffffffff” that actually caused Windows to take an extra five seconds to boot.
If you’re a fan of Microsoft’s automated “Fix It” apps, there are several options to disable or re-enable certain or all IPv6 components (which Microsoft doesn’t recommend) and you can get those here.
If you would rather do it yourself manually:
- Open the registry editor as a member of the local admin group and navigate to:
HKLM\SYSTEM\CurrentControlSet\Services\TCPIP6\Parameters
- Right-click on Parameters, add a new DWORD (32 Bit) Value, and name it “DisabledComponents”.
- Double-click on DisabledComponents to modify it, set the Base to hexadecimal, and enter one of the values below:
To disable all IPv6 components: "ff" To prefer IPv4 over IPv6 in prefix policies: "20" To disable IPv6 on all nontunnel interfaces: "10" To disable IPv6 on all tunnel interfaces: "1" To disable IPv6 on nontunnel interfaces (except the loopback) and on IPv6 tunnel interface: "11"
- Reboot
I’ve come to the conclusion that the problem described above with the Cisco VPN client appears to be some sort of bug that happens randomly and since it’s not supported any more, the only real fix is to use Shrewsoft VPN client instead as it does not have this issue.