Mirror/k3s
1
0
Fork 0
mirror of https://github.com/k3s-io/k3s.git synced 2024-06-07 19:41:36 +00:00
Code Issues Packages Projects Releases Wiki Activity

Allow for multiple etcd snapshot restoration (#2307)

Browse Source 
* add reset tmp file

Signed-off-by: galal-hussein <hussein.galal.ahmed.11@gmail.com>

* go imports

Signed-off-by: galal-hussein <hussein.galal.ahmed.11@gmail.com>

* fix multiple lines string

Signed-off-by: galal-hussein <hussein.galal.ahmed.11@gmail.com>

* fix typo

Signed-off-by: galal-hussein <hussein.galal.ahmed.11@gmail.com>

* use resetFile function

Signed-off-by: galal-hussein <hussein.galal.ahmed.11@gmail.com>
This commit is contained in:
Hussein Galal 2020-09-30 02:53:31 +02:00 committed by GitHub
parent 8262e23169
commit 373449ec0a
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 46 additions and 29 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
pkg/cluster/managed.go
View File

@ -5,11 +5,16 @@ package cluster
import (
"context"
"fmt"
"net/http"
"os"
"strings"
"time"
"github.com/rancher/k3s/pkg/etcd"
"github.com/rancher/k3s/pkg/cluster/managed"
"github.com/rancher/k3s/pkg/version"
"github.com/rancher/kine/pkg/endpoint"
"github.com/sirupsen/logrus"
)
@ -47,13 +52,23 @@ func (c *Cluster) testClusterDB(ctx context.Context) (<-chan struct{}, error) {
// start starts the database, unless a cluster reset has been requested, in which case
// it does that instead.
func (c *Cluster) start(ctx context.Context) error {
resetFile := etcd.ResetFile(c.config)
if c.managedDB == nil {
return nil
}
if c.config.ClusterReset {
if _, err := os.Stat(resetFile); err != nil {
if !os.IsNotExist(err) {
return err
}
} else {
return fmt.Errorf("cluster-reset was successfully performed, please remove the cluster-reset flag and start %s normally, if you need to perform another cluster reset, you must first manually delete the %s file", version.Program, resetFile)
}
return c.managedDB.Reset(ctx, c.clientAccessInfo)
}
// removing the reset file and ignore error if the file doesnt exist
os.Remove(resetFile)
return c.managedDB.Start(ctx, c.clientAccessInfo)
}

60
pkg/etcd/etcd.go
View File

@ -120,6 +120,11 @@ func nameFile(config *config.Control) string {
return filepath.Join(etcdDBDir(config), "name")
}
// ResetFile returns the path to etcdDBDir/reset-flag
func ResetFile(config *config.Control) string {
return filepath.Join(config.DataDir, "db", "reset-flag")
}
// IsInitialized checks to see if a WAL directory exists. If so, we assume that etcd
// has already been brought up at least once.
func (e *ETCD) IsInitialized(ctx context.Context, config *config.Control) (bool, error) {
@ -171,7 +176,10 @@ func (e *ETCD) Reset(ctx context.Context, clientAccessInfo *clientaccess.Info) e
if err := e.setName(true); err != nil {
return err
}
// touch a file to avoid multiple resets
if err := ioutil.WriteFile(ResetFile(e.config), []byte{}, 0600); err != nil {
return err
}
return e.newCluster(ctx, true)
}
@ -632,34 +640,28 @@ func (e *ETCD) setSnapshotFunction(ctx context.Context) {
// completion.
func (e *ETCD) Restore(ctx context.Context) error {
// check the old etcd data dir
oldDataDir := etcdDBDir(e.config) + "-old"
if s, err := os.Stat(oldDataDir); err == nil && s.IsDir() {
logrus.Infof("Etcd already restored from a snapshot. Restart without --snapshot-restore-path flag. Backup and delete ${datadir}/server/db on each peer etcd server and rejoin the nodes")
os.Exit(0)
} else if os.IsNotExist(err) {
if e.config.ClusterResetRestorePath == "" {
return errors.New("no etcd restore path was specified")
}
// make sure snapshot exists before restoration
if _, err := os.Stat(e.config.ClusterResetRestorePath); err != nil {
return err
}
// move the data directory to a temp path
if err := os.Rename(etcdDBDir(e.config), oldDataDir); err != nil {
return err
}
sManager := snapshot.NewV3(nil)
if err := sManager.Restore(snapshot.RestoreConfig{
SnapshotPath: e.config.ClusterResetRestorePath,
Name: e.name,
OutputDataDir: etcdDBDir(e.config),
OutputWALDir: walDir(e.config),
PeerURLs: []string{e.peerURL()},
InitialCluster: e.name + "=" + e.peerURL(),
}); err != nil {
return err
}
} else {
oldDataDir := etcdDBDir(e.config) + "-old-" + strconv.Itoa(int(time.Now().Unix()))
if e.config.ClusterResetRestorePath == "" {
return errors.New("no etcd restore path was specified")
}
// make sure snapshot exists before restoration
if _, err := os.Stat(e.config.ClusterResetRestorePath); err != nil {
return err
}
// move the data directory to a temp path
if err := os.Rename(etcdDBDir(e.config), oldDataDir); err != nil {
return err
}
logrus.Infof("Pre-restore etcd database moved to %s", oldDataDir)
sManager := snapshot.NewV3(nil)
if err := sManager.Restore(snapshot.RestoreConfig{
SnapshotPath: e.config.ClusterResetRestorePath,
Name: e.name,
OutputDataDir: etcdDBDir(e.config),
OutputWALDir: walDir(e.config),
PeerURLs: []string{e.peerURL()},
InitialCluster: e.name + "=" + e.peerURL(),
}); err != nil {
return err
}
return nil