Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Shortening datapath downtime for libreswan cable
Currently there is about 2 seconds of downtime in datapath for libreswan cable when the gateway pod restarts. When Pluto starts running [1] the IPsec datapath rules created by the previous GW pod are cleared, and traffic is only recovered after ConnectToEndpoint events are handled by the libreswan cable driver. This PR reduces datapath downtime by: A. Checking Pluto control socket status at faster rate B. Starting Pluto right before the first ConnectToEndpoint event is handled by libreswan cable driver. With this PR, the datapath downtime is reduced to ~0.5 seconds. [1] https://github.com/submariner-io/submariner/blob/devel/pkg/cable/libreswan/libreswan.go#L158 Signed-off-by: Yossi Boaron <yboaron@redhat.com>
- Loading branch information