Files
my-rook-config/pkg/operator/ceph/nfs/nfs.go
T
Sébastien Han b89730d895 ceph: refactor operator initialization sequence
This commit is a large refactor on how the operator starts, stops and
how it starts various sub-components such as the ceph-csi driver. It
also refines the way we cancel orchestrations. We don't use breakpoints
anymore but send our self a SIGUP to reload our controller runtime
manager.
The reload will happen under different circonstances like:

* a new adminission controller secret is created/deleted/changed
* a CephCluster CR is edited

As mentioned earlier, the csi driver now has its own controller, just
like flex. It reacts to change in the operator config map for particular
ROOK_CSI_ fields.

A second new controller for the operator's general config has been
created, it manages:

* the logging level
* the ceph CLI command timeout
* the discovery daemon

The operator reacts much more rapidly to cancellation events by stopping
the manager's context and reloading it.

Signed-off-by: Sébastien Han <seb@redhat.com>
2021-09-17 16:57:12 +02:00

280 lines
9.0 KiB
Go

/*
Copyright 2018 The Rook Authors. All rights reserved.
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 nfs manages NFS ganesha servers for Ceph
package nfs
import (
"fmt"
"github.com/banzaicloud/k8s-objectmatcher/patch"
"github.com/pkg/errors"
cephv1 "github.com/rook/rook/pkg/apis/ceph.rook.io/v1"
"github.com/rook/rook/pkg/clusterd"
cephclient "github.com/rook/rook/pkg/daemon/ceph/client"
opmon "github.com/rook/rook/pkg/operator/ceph/cluster/mon"
"github.com/rook/rook/pkg/operator/ceph/config"
"github.com/rook/rook/pkg/operator/k8sutil"
v1 "k8s.io/api/core/v1"
kerrors "k8s.io/apimachinery/pkg/api/errors"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"sigs.k8s.io/controller-runtime/pkg/controller/controllerutil"
)
const (
ganeshaRadosGraceCmd = "ganesha-rados-grace"
)
var updateDeploymentAndWait = opmon.UpdateCephDeploymentAndWait
type daemonConfig struct {
ID string // letter ID of daemon (e.g., a, b, c, ...)
ConfigConfigMap string // name of configmap holding config
DataPathMap *config.DataPathMap // location to store data in container
}
// Create the ganesha server
func (r *ReconcileCephNFS) upCephNFS(n *cephv1.CephNFS) error {
for i := 0; i < n.Spec.Server.Active; i++ {
id := k8sutil.IndexToName(i)
configName, err := r.createConfigMap(n, id)
if err != nil {
return errors.Wrap(err, "failed to create config")
}
err = r.addRADOSConfigFile(n, id)
if err != nil {
return errors.Wrap(err, "failed to create RADOS config object")
}
cfg := daemonConfig{
ID: id,
ConfigConfigMap: configName,
DataPathMap: &config.DataPathMap{
HostDataDir: "", // nfs daemon does not store data on host, ...
ContainerDataDir: cephclient.DefaultConfigDir, // does share data in containers using emptyDir, ...
HostLogAndCrashDir: "", // and does not log to /var/log/ceph dir
},
}
err = r.generateKeyring(n, id)
if err != nil {
return errors.Wrapf(err, "failed to generate keyring for %q", id)
}
// create the deployment
deployment, err := r.makeDeployment(n, cfg)
if err != nil {
return errors.Wrap(err, "failed to set to create deployment")
}
// Set owner ref to cephNFS object
err = controllerutil.SetControllerReference(n, deployment, r.scheme)
if err != nil {
return errors.Wrapf(err, "failed to set owner reference for ceph nfs deployment %q", deployment.Name)
}
// Set the deployment hash as an annotation
err = patch.DefaultAnnotator.SetLastAppliedAnnotation(deployment)
if err != nil {
return errors.Wrapf(err, "failed to set annotation for deployment %q", deployment.Name)
}
// start the deployment
_, err = r.context.Clientset.AppsV1().Deployments(n.Namespace).Create(r.opManagerContext, deployment, metav1.CreateOptions{})
if err != nil {
if !kerrors.IsAlreadyExists(err) {
return errors.Wrap(err, "failed to create ceph nfs deployment")
}
logger.Infof("ceph nfs deployment %q already exists. updating if needed", deployment.Name)
if err := updateDeploymentAndWait(r.context, r.clusterInfo, deployment, "nfs", id, r.cephClusterSpec.SkipUpgradeChecks, false); err != nil {
return errors.Wrapf(err, "failed to update ceph nfs deployment %q", deployment.Name)
}
} else {
logger.Infof("ceph nfs deployment %q started", deployment.Name)
}
// create a service
err = r.createCephNFSService(n, cfg)
if err != nil {
return errors.Wrap(err, "failed to create ceph nfs service")
}
// Add server to database
err = r.addServerToDatabase(n, id)
if err != nil {
return errors.Wrapf(err, "failed to add server %q to database", id)
}
}
return nil
}
// Create empty config file for new ganesha server
func (r *ReconcileCephNFS) addRADOSConfigFile(n *cephv1.CephNFS, name string) error {
config := getGaneshaConfigObject(n, r.clusterInfo.CephVersion, name)
cmd := "rados"
args := []string{
"--pool", n.Spec.RADOS.Pool,
"--namespace", n.Spec.RADOS.Namespace,
"--conf", cephclient.CephConfFilePath(r.context.ConfigDir, n.Namespace),
}
err := r.context.Executor.ExecuteCommand(cmd, append(args, "stat", config)...)
if err == nil {
// If stat works then we assume it's present already
return nil
}
// try to create it
return r.context.Executor.ExecuteCommand(cmd, append(args, "create", config)...)
}
func (r *ReconcileCephNFS) addServerToDatabase(nfs *cephv1.CephNFS, name string) error {
logger.Infof("adding ganesha %q to grace db", name)
if err := r.runGaneshaRadosGrace(nfs, name, "add"); err != nil {
return errors.Wrapf(err, "failed to add %q to grace db", name)
}
return nil
}
func (r *ReconcileCephNFS) removeServerFromDatabase(nfs *cephv1.CephNFS, name string) {
logger.Infof("removing ganesha %q from grace db", name)
if err := r.runGaneshaRadosGrace(nfs, name, "remove"); err != nil {
logger.Errorf("failed to remove %q from grace db. %v", name, err)
}
}
func (r *ReconcileCephNFS) runGaneshaRadosGrace(nfs *cephv1.CephNFS, name, action string) error {
nodeID := getNFSNodeID(nfs, name)
cmd := ganeshaRadosGraceCmd
args := []string{"--pool", nfs.Spec.RADOS.Pool, "--ns", nfs.Spec.RADOS.Namespace, action, nodeID}
env := []string{fmt.Sprintf("CEPH_CONF=%s", cephclient.CephConfFilePath(r.context.ConfigDir, nfs.Namespace))}
return r.context.Executor.ExecuteCommandWithEnv(env, cmd, args...)
}
func (r *ReconcileCephNFS) generateConfigMap(n *cephv1.CephNFS, name string) *v1.ConfigMap {
data := map[string]string{
"config": getGaneshaConfig(n, r.clusterInfo.CephVersion, name),
}
configMap := &v1.ConfigMap{
ObjectMeta: metav1.ObjectMeta{
Name: instanceName(n, name),
Namespace: n.Namespace,
Labels: getLabels(n, name, true),
},
Data: data,
}
return configMap
}
func (r *ReconcileCephNFS) createConfigMap(n *cephv1.CephNFS, name string) (string, error) {
// Generate configMap
configMap := r.generateConfigMap(n, name)
// Set owner reference
err := controllerutil.SetControllerReference(n, configMap, r.scheme)
if err != nil {
return "", errors.Wrapf(err, "failed to set owner reference for ceph ganesha configmap %q", configMap.Name)
}
if _, err := r.context.Clientset.CoreV1().ConfigMaps(n.Namespace).Create(r.opManagerContext, configMap, metav1.CreateOptions{}); err != nil {
if !kerrors.IsAlreadyExists(err) {
return "", errors.Wrap(err, "failed to create ganesha config map")
}
logger.Debugf("updating config map %q that already exists", configMap.Name)
if _, err = r.context.Clientset.CoreV1().ConfigMaps(n.Namespace).Update(r.opManagerContext, configMap, metav1.UpdateOptions{}); err != nil {
return "", errors.Wrap(err, "failed to update ganesha config map")
}
}
return configMap.Name, nil
}
// Down scale the ganesha server
func (r *ReconcileCephNFS) downCephNFS(n *cephv1.CephNFS, nfsServerListNum int) error {
diffCount := nfsServerListNum - n.Spec.Server.Active
for i := 0; i < diffCount; {
depIDToRemove := nfsServerListNum - 1
name := k8sutil.IndexToName(depIDToRemove)
depNameToRemove := instanceName(n, name)
// Remove deployment
logger.Infof("removing deployment %q", depNameToRemove)
err := r.context.Clientset.AppsV1().Deployments(n.Namespace).Delete(r.opManagerContext, depNameToRemove, metav1.DeleteOptions{})
if err != nil {
if !kerrors.IsNotFound(err) {
return errors.Wrap(err, "failed to delete ceph nfs deployment")
}
}
// Remove from grace db
r.removeServerFromDatabase(n, name)
nfsServerListNum = nfsServerListNum - 1
i++
}
return nil
}
func (r *ReconcileCephNFS) removeServersFromDatabase(n *cephv1.CephNFS, newActive int) error {
for i := n.Spec.Server.Active - 1; i >= newActive; i-- {
name := k8sutil.IndexToName(i)
r.removeServerFromDatabase(n, name)
}
return nil
}
func instanceName(n *cephv1.CephNFS, name string) string {
return fmt.Sprintf("%s-%s-%s", AppName, n.Name, name)
}
func validateGanesha(context *clusterd.Context, clusterInfo *cephclient.ClusterInfo, n *cephv1.CephNFS) error {
// core properties
if n.Name == "" {
return errors.New("missing name")
}
if n.Namespace == "" {
return errors.New("missing namespace")
}
// Client recovery properties
if n.Spec.RADOS.Pool == "" {
return errors.New("missing RADOS.pool")
}
// Ganesha server properties
if n.Spec.Server.Active == 0 {
return errors.New("at least one active server required")
}
// The existence of the pool provided in n.Spec.RADOS.Pool is necessary otherwise addRADOSConfigFile() will fail
_, err := cephclient.GetPoolDetails(context, clusterInfo, n.Spec.RADOS.Pool)
if err != nil {
return errors.Wrapf(err, "pool %q not found", n.Spec.RADOS.Pool)
}
return nil
}