# worker node disruptions

- [Ungraceful node termination leads to 15mins of failing requests #33466](https://github.com/istio/istio/issues/33466)
- [If a node terminates ungracefully, sidecars route to bad pods #28865](https://github.com/istio/istio/issues/28865)
- [Add support for TCP_USER_TIMEOUT socket options for envoy clusters #38476](https://github.com/istio/istio/issues/38476)
-