Better isolates the K3s implementation from the interface, and aligns
the package path with other projects executors. This should also remove
the indirect flannel dep from other projects that don't use the embedded
executor.
Signed-off-by: Brad Davidson <brad.davidson@rancher.com>
Allows properly delegating CNI startup to executor, so that it can be plugged in as platform and distro specific implimentation without relying on cli flag hacks
Signed-off-by: Brad Davidson <brad.davidson@rancher.com>
Fixes issue where the apiserver on control-plane-only nodes does not
actually wait for a connection to etcd to be available before starting.
Signed-off-by: Brad Davidson <brad.davidson@rancher.com>
Currently only waits on etcd and kine, as other components
are stateless and do not need to shut down cleanly.
Terminal but non-fatal errors now request shutdown via context
cancellation, instead of just logging a fatal error.
Signed-off-by: Brad Davidson <brad.davidson@rancher.com>
Addresses flakes in etcd CI due to the port still being in TIME_WAIT after the server is shut down between tests
Signed-off-by: Brad Davidson <brad.davidson@rancher.com>
This eliminates the final channel that was being passed around in an internal struct. The ETCD management code passes in a func that can be polled until etcd is ready; the executor is responsible for polling this after etcd is started and closing the etcd ready channel at the correct time.
Signed-off-by: Brad Davidson <brad.davidson@rancher.com>
Move the container runtime ready channel into the executor interface, instead of passing it awkwardly between server and agent config structs
Signed-off-by: Brad Davidson <brad.davidson@rancher.com>
Splits server startup into prepare/start phases. Server's agent is now
started after server is prepared, but before it is started. This allows
us to properly bootstrap the executor before starting server components,
and use the executor to provide a shared channel to wait on apiserver
readiness.
This allows us to replace four separate callers of WaitForAPIServerReady
with reads from a common ready channel.
Signed-off-by: Brad Davidson <brad.davidson@rancher.com>
* Bump etcd to v3.5.4-k3s1
* Fix issue with datastore corruption on cluster-reset
* Disable unnecessary components during cluster reset
Disable control-plane components and the tunnel setup during
cluster-reset, even when not doing a restore. This reduces the amount of
log clutter during cluster reset/restore, making any errors encountered
more obvious.
Signed-off-by: Brad Davidson <brad.davidson@rancher.com>
* Add etcd extra args support for K3s
Signed-off-by: Chris Kim <oats87g@gmail.com>
* Add etcd custom argument integration test
Signed-off-by: Chris Kim <oats87g@gmail.com>
* go generate
Signed-off-by: Chris Kim <oats87g@gmail.com>
* Move cloud-controller-manager into an embedded executor
* Import K3s cloud provider and clean up imports
Signed-off-by: Chris Kim <oats87g@gmail.com>
* Set etcd timeouts using values from k8s instead of etcdctl
Fix for one of the warnings from #2303
* Use etcd zap logger instead of deprecated capsnlog
Fix for one of the warnings from #2303
* Remove member self-promotion code paths
* Add learner promotion tracking code
* Fix RaftAppliedIndex progress check
* Remove ErrGRPCKeyNotFound check
This is not used by v3 API - it just returns a response with 0 KVs.
Signed-off-by: Brad Davidson <brad.davidson@rancher.com>
In rke2 everything is a static pod so this causes a chicken and egg situation
in which we need the kubelet running before the kube-apiserver can be
launched. By starting the apiserver in the background this allows us to
do this odd bootstrapping.
