KeepAliveTime and KeepAliveInterval parameters, you can control how often the system sends probes to maintain idle connections. This process ensures that stale sockets are closed correctly, which is vital for applications requiring persistent network links or high concurrency.
Understanding TCP Keepalive Behavior
By default, Windows uses a TCP keepalive interval of two hours (7,200,000 milliseconds). This long duration means that if a connection becomes inactive due to a dropped link or firewall timeout, the operating system will not detect it for a very long time. Consequently, applications may hang waiting for data that will never arrive because the remote end has already disconnected.
Enabling and tuning these settings allows your machine to detect dead connections much faster. The primary keys involved in this configuration are located within the TCP/IP parameters registry hive. Adjusting these values helps prevent issues where applications appear frozen or fail to reconnect automatically after a network interruption.
Locating the Registry Keys
To modify the TCP keepalive settings, you must access the Windows Registry Editor. It is crucial to create a backup before making any changes to ensure you can restore previous settings if needed. Open the Run dialog by pressing Win + R, type regedit, and press Enter. Navigate through the following path:
HKEY_LOCAL_MACHINESYSTEMCurrentControlSetServicesTcpipParameters
Once you are in the Parameters folder, look at the right-hand pane. You will see several DWORD (32-bit) values related to TCP behavior. If the specific keepalive keys do not exist, you may need to create them manually, though they are typically present on modern Windows 10 installations.
Configuring KeepAliveTime
The most critical setting is KeepAliveTime. This value determines the total time, in milliseconds, that a connection must remain idle before TCP starts sending keepalive messages. The default value is usually set to 7,200,000 ms (two hours). To enable more aggressive keepalive probing, you should change this value to a lower number, such as 300,000 ms (five minutes).
Double-click on KeepAliveTime and select Decimal as the base. Enter your desired time in milliseconds. For example, entering 300000 will cause the system to start checking the connection health every five minutes of inactivity. After changing this value, click OK to save it. This single change significantly reduces the time it takes to detect broken pipes.
Adjusting KeepAliveInterval
While KeepAliveTime sets the delay before probing begins, KeepAliveInterval controls how frequently the probes are sent once the timer expires. The default interval is typically 1 second (1,000 ms). If you want the system to send multiple rapid probes to confirm connectivity, you can reduce this interval to 500 ms or even less. However, reducing it too much may generate unnecessary network traffic.
Modify KeepAliveInterval similarly to the previous step. Set it to a decimal value representing the milliseconds between each probe. A common configuration is a KeepAliveTime of 300,000 ms and an KeepAliveInterval of 1,000 ms. This combination balances network efficiency with reliable connection detection.
Applying Changes and Verification
Registry changes take effect immediately for new connections; existing active connections may continue using the old parameters until they close. Therefore, restarting the affected applications or rebooting the computer is recommended to ensure all services adopt the new settings. After applying the changes, you can verify the configuration by returning to the registry editor and confirming the decimal values match your input.
If you encounter issues with other network configurations, such as high concurrency limits, refer to our guide on resolving Windows can't connect more than 64k concurrency outgoing connection errors. These settings work in tandem to optimize overall network stability.
For users managing complex network setups, understanding how interfaces interact is also important. You might find it useful to read about Windows Dual NIC, bridge interfaces to ensure your keepalive settings apply correctly across different network segments.
Frequently Asked Questions
Does enabling TCP keepalive increase network traffic?
Yes, but the impact is negligible for most desktop and server applications. The keepalive packets are very small and only sent during periods of inactivity, so they do not consume significant bandwidth during normal data transfer operations.
What happens if I set KeepAliveTime to zero?
Setting the value to zero disables TCP keepalive entirely. This reverts the system to its default behavior of ignoring idle connection states unless the application itself implements a higher-level heartbeat mechanism, potentially leading to hung connections.
Can I use netsh to change these settings?
No, the netsh command does not support modifying TCP keepalive timers. These values are strictly controlled via the Windows Registry under the Tcpip Parameters key, as described in the steps above.
Will this fix slow internet speeds?
No, TCP keepalive is designed to manage connection state, not throughput. If you are experiencing slow speeds, consider checking for background updates or consulting resources on How to Fix Slow Internet Connection Due to Second Local Network issues.
Are there risks associated with lowering these intervals?
The primary risk is increased CPU usage on extremely high-volume servers due to more frequent packet processing. For standard client machines, the performance impact is unnoticeable, and the benefit of detecting dead connections outweighs the minor overhead.