forked from crewjam/etcd-aws
-
Notifications
You must be signed in to change notification settings - Fork 0
/
backup.go
252 lines (224 loc) · 6.78 KB
/
backup.go
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
package main
import (
"compress/gzip"
"encoding/json"
"fmt"
"io"
"net/http"
"time"
log "github.com/Sirupsen/logrus"
"github.com/aws/aws-sdk-go/aws"
"github.com/aws/aws-sdk-go/aws/awserr"
"github.com/aws/aws-sdk-go/service/cloudwatch"
"github.com/aws/aws-sdk-go/service/ec2"
"github.com/aws/aws-sdk-go/service/s3"
"github.com/aws/aws-sdk-go/service/s3/s3manager"
"github.com/coreos/go-etcd/etcd"
"github.com/crewjam/ec2cluster"
)
// backupService invokes backupOnce() periodically if the current node is the cluster leader.
func backupService(s *ec2cluster.Cluster, backupBucket, backupKey, dataDir string, interval time.Duration) error {
instance, err := s.Instance()
if err != nil {
return err
}
ticker := time.Tick(interval)
for {
<-ticker
resp, err := http.Get(fmt.Sprintf("http://%s:2379/v2/stats/self", *instance.PrivateIpAddress))
if err != nil {
return fmt.Errorf("%s: http://%s:2379/v2/stats/self: %s", *instance.InstanceId,
*instance.PrivateIpAddress, err)
}
nodeState := etcdState{}
if err := json.NewDecoder(resp.Body).Decode(&nodeState); err != nil {
return fmt.Errorf("%s: http://%s:2379/v2/stats/self: %s", *instance.InstanceId,
*instance.PrivateIpAddress, err)
}
// if the cluster has a leader other than the current node, then don't do the
// backup.
if nodeState.LeaderInfo.Leader != "" && nodeState.ID != nodeState.LeaderInfo.Leader {
log.Printf("backup: %s: http://%s:2379/v2/stats/self: not the leader", *instance.InstanceId,
*instance.PrivateIpAddress)
<-ticker
continue
}
if err := backupOnce(s, backupBucket, backupKey, dataDir); err != nil {
return err
}
}
panic("not reached")
}
// getInstanceTag returns the first occurrence of the specified tag on the instance
// or an empty string if the tag is not found.
func getInstanceTag(instance *ec2.Instance, tagName string) string {
rv := ""
for _, tag := range instance.Tags {
if *tag.Key == tagName {
rv = *tag.Value
}
}
return rv
}
// dumpEtcdNode writes a JSON representation of the nodes and and below `key`
// to `w`. Returns the number of nodes traversed.
func dumpEtcdNode(key string, etcdClient *etcd.Client, w io.Writer) (int, error) {
response, err := etcdClient.Get(key, false, false)
if err != nil {
return 0, err
}
childNodes := response.Node.Nodes
response.Node.Nodes = nil
if err := json.NewEncoder(w).Encode(response.Node); err != nil {
return 0, err
}
count := 1
// enumerate all the child nodes. If a child node is
// a directory, then it must be backed up recursively.
// Otherwise it can just be written
for _, childNode := range childNodes {
if childNode.Dir {
c, err := dumpEtcdNode(childNode.Key, etcdClient, w)
if err != nil {
return 0, err
}
count += c
} else {
if err := json.NewEncoder(w).Encode(childNode); err != nil {
return 0, err
}
count += 1
}
}
return count, nil
}
// backupOnce dumps all the nodes in the etcd cluster to the specified S3 bucket. On
// success it emits a CloudWatch metric for the number of keys backed up. The absence
// of data on this metric indicates the backup has failed.
func backupOnce(s *ec2cluster.Cluster, backupBucket, backupKey, dataDir string) error {
instance, err := s.Instance()
if err != nil {
return err
}
etcdClient := etcd.NewClient([]string{fmt.Sprintf("http://%s:2379", *instance.PrivateIpAddress)})
if success := etcdClient.SyncCluster(); !success {
return fmt.Errorf("backupOnce: cannot sync machines")
}
var valueCount int
bodyReader, bodyWriter := io.Pipe()
go func() {
gzWriter := gzip.NewWriter(bodyWriter)
var err error
valueCount, err = dumpEtcdNode("/", etcdClient, gzWriter)
if err != nil {
gzWriter.Close()
bodyWriter.CloseWithError(err)
return
}
gzWriter.Close()
bodyWriter.Close()
}()
s3mgr := s3manager.NewUploader(s.AwsSession)
_, err = s3mgr.Upload(&s3manager.UploadInput{
Key: aws.String(backupKey),
Bucket: aws.String(backupBucket),
Body: bodyReader,
ServerSideEncryption: aws.String(s3.ServerSideEncryptionAes256),
ACL: aws.String(s3.ObjectCannedACLPrivate),
})
if err != nil {
return fmt.Errorf("upload s3://%s%s: %s", backupBucket, backupKey, err)
}
log.Printf("backup written to s3://%s%s (%d values)", backupBucket, backupKey,
valueCount)
cloudwatch.New(s.AwsSession).PutMetricData(&cloudwatch.PutMetricDataInput{
Namespace: aws.String("Local/etcd"),
MetricData: []*cloudwatch.MetricDatum{
&cloudwatch.MetricDatum{
MetricName: aws.String("BackupKeyCount"),
Dimensions: []*cloudwatch.Dimension{
&cloudwatch.Dimension{
Name: aws.String("AutoScalingGroupName"),
Value: aws.String(getInstanceTag(instance, "aws:autoscaling:groupName")),
},
},
Unit: aws.String(cloudwatch.StandardUnitCount),
Value: aws.Float64(float64(valueCount)),
},
},
})
return nil
}
// loadEtcdNode reads `r` containing JSON objects representing etcd nodes and
// loads them into server.
func loadEtcdNode(etcdClient *etcd.Client, r io.Reader) error {
jsonReader := json.NewDecoder(r)
for {
node := etcd.Node{}
if err := jsonReader.Decode(&node); err != nil {
if err == io.EOF {
break
}
return err
}
if node.Key == "" && node.Dir {
continue // skip root
}
// compute a new TTL
ttl := 0
if node.Expiration != nil {
ttl := node.Expiration.Sub(time.Now()).Seconds()
if ttl < 0 {
// expired, skip
continue
}
}
if node.Dir {
_, err := etcdClient.SetDir(node.Key, uint64(ttl))
if err != nil {
return fmt.Errorf("%s: %s", node.Key, err)
}
} else {
_, err := etcdClient.Set(node.Key, node.Value, uint64(ttl))
if err != nil {
return fmt.Errorf("%s: %s", node.Key, err)
}
}
}
return nil
}
// restoreBackup reads the backup from S3 and applies it to the current cluster.
func restoreBackup(s *ec2cluster.Cluster, backupBucket, backupKey, dataDir string) error {
instance, err := s.Instance()
if err != nil {
return err
}
etcdClient := etcd.NewClient([]string{fmt.Sprintf("http://%s:2379", *instance.PrivateIpAddress)})
if success := etcdClient.SyncCluster(); !success {
return fmt.Errorf("restore: cannot sync machines")
}
s3svc := s3.New(s.AwsSession)
resp, err := s3svc.GetObject(&s3.GetObjectInput{
Key: &backupKey,
Bucket: &backupBucket,
})
if err != nil {
if reqErr, ok := err.(awserr.RequestFailure); ok {
if reqErr.StatusCode() == http.StatusNotFound {
log.Printf("restore: s3://%s%s does not exist", backupBucket, backupKey)
return nil
}
}
return fmt.Errorf("cannot fetch backup file: %s", err)
}
gzipReader, err := gzip.NewReader(resp.Body)
if err != nil {
return err
}
defer gzipReader.Close()
if err := loadEtcdNode(etcdClient, gzipReader); err != nil {
return err
}
log.Printf("restore: complete")
return nil
}