Our organization recently updated all of our devices to WARP Client 2023.7.160.0. Ever since then we are seeing random CPU spikes to 100%. During this spike we are experiencing massive network lag.
Terminating the client and service resolves the issue indefinitely.
I am about to remove this version and reinstall an older version as a test.
Continued testing today. The below highlighted items are spiking to 100% CPU very briefly while the WARP client is running. This only happens with the WARP client running. Tested multiple version of the client with the same behavior.
So this has also been an issue for users of Cisco Umbrella clients. OpenDNS/Cisco support has found a work around:
Disable IP Layer enforcement
Well this is not an option that I see in WARP settings in the portal or on the client.
Whether or not the issue is even remotely the same, there is one thing for sure, WARP agent is causing the same behavior in the “LocalServiceNoNetworkFirewall” service.
“We found that our IP layer enforcement routine was causing high CPU usage in the way it utilized the Base Filtering Engine (BFE).”
“The BFE service tends to churn CPU when IP firewall rules are added or removed. So if Avast! is using it for blocking bad IPs, that’s likely what’s happening. It may be loading and unloading the rules too often?”
Sounds like WARP may be using the BFE service as well.
Installing a 3rd party firewall such as BitDefender’s firewall allows the WARP client to perform better.
Seems as though the Windows Defender Firewall (BFE) performs poorly in situations where additional rules and policies are being created/managed/used by other software such as the WARP client.
I know there is a documented issue with WARP and the split tunnel records causing performance issues. Is this because WARP is using the Windows Defender Firewall BFE?
