5614a9d064
The "todo" packages were necessary while moving code around to avoid hitting cyclic dependencies. Now that any sub package can depend on the framework, they are no longer needed and the code can be moved into the normal sub packages.
94 lines
2.6 KiB
Go
94 lines
2.6 KiB
Go
/*
|
|
Copyright 2014 The Kubernetes Authors.
|
|
|
|
Licensed under the Apache License, Version 2.0 (the "License");
|
|
you may not use this file except in compliance with the License.
|
|
You may obtain a copy of the License at
|
|
|
|
http://www.apache.org/licenses/LICENSE-2.0
|
|
|
|
Unless required by applicable law or agreed to in writing, software
|
|
distributed under the License is distributed on an "AS IS" BASIS,
|
|
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
See the License for the specific language governing permissions and
|
|
limitations under the License.
|
|
*/
|
|
|
|
package node
|
|
|
|
import (
|
|
"sync"
|
|
"time"
|
|
|
|
"github.com/onsi/ginkgo/v2"
|
|
|
|
v1 "k8s.io/api/core/v1"
|
|
"k8s.io/apimachinery/pkg/util/wait"
|
|
clientset "k8s.io/client-go/kubernetes"
|
|
|
|
"k8s.io/kubernetes/test/e2e/framework"
|
|
e2essh "k8s.io/kubernetes/test/e2e/framework/ssh"
|
|
)
|
|
|
|
// NodeKiller is a utility to simulate node failures.
|
|
type NodeKiller struct {
|
|
config framework.NodeKillerConfig
|
|
client clientset.Interface
|
|
provider string
|
|
}
|
|
|
|
// NewNodeKiller creates new NodeKiller.
|
|
func NewNodeKiller(config framework.NodeKillerConfig, client clientset.Interface, provider string) *NodeKiller {
|
|
config.NodeKillerStopCh = make(chan struct{})
|
|
return &NodeKiller{config, client, provider}
|
|
}
|
|
|
|
// Run starts NodeKiller until stopCh is closed.
|
|
func (k *NodeKiller) Run(stopCh <-chan struct{}) {
|
|
// wait.JitterUntil starts work immediately, so wait first.
|
|
time.Sleep(wait.Jitter(k.config.Interval, k.config.JitterFactor))
|
|
wait.JitterUntil(func() {
|
|
nodes := k.pickNodes()
|
|
k.kill(nodes)
|
|
}, k.config.Interval, k.config.JitterFactor, true, stopCh)
|
|
}
|
|
|
|
func (k *NodeKiller) pickNodes() []v1.Node {
|
|
nodes, err := GetReadySchedulableNodes(k.client)
|
|
framework.ExpectNoError(err)
|
|
numNodes := int(k.config.FailureRatio * float64(len(nodes.Items)))
|
|
|
|
nodes, err = GetBoundedReadySchedulableNodes(k.client, numNodes)
|
|
framework.ExpectNoError(err)
|
|
return nodes.Items
|
|
}
|
|
|
|
func (k *NodeKiller) kill(nodes []v1.Node) {
|
|
wg := sync.WaitGroup{}
|
|
wg.Add(len(nodes))
|
|
for _, node := range nodes {
|
|
node := node
|
|
go func() {
|
|
defer ginkgo.GinkgoRecover()
|
|
defer wg.Done()
|
|
|
|
framework.Logf("Stopping docker and kubelet on %q to simulate failure", node.Name)
|
|
err := e2essh.IssueSSHCommand("sudo systemctl stop docker kubelet", k.provider, &node)
|
|
if err != nil {
|
|
framework.Logf("ERROR while stopping node %q: %v", node.Name, err)
|
|
return
|
|
}
|
|
|
|
time.Sleep(k.config.SimulatedDowntime)
|
|
|
|
framework.Logf("Rebooting %q to repair the node", node.Name)
|
|
err = e2essh.IssueSSHCommand("sudo reboot", k.provider, &node)
|
|
if err != nil {
|
|
framework.Logf("ERROR while rebooting node %q: %v", node.Name, err)
|
|
return
|
|
}
|
|
}()
|
|
}
|
|
wg.Wait()
|
|
}
|