-
Notifications
You must be signed in to change notification settings - Fork 35
/
cleanup.go
177 lines (147 loc) · 5.04 KB
/
cleanup.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
package evacuation
import (
"errors"
"fmt"
"os"
"sync"
"time"
"code.cloudfoundry.org/bbs"
"code.cloudfoundry.org/bbs/models"
"code.cloudfoundry.org/clock"
loggingclient "code.cloudfoundry.org/diego-logging-client"
"code.cloudfoundry.org/executor"
"code.cloudfoundry.org/lager/v3"
)
const (
exitTimeoutOffset = 5 * time.Second
)
var strandedEvacuatingActualLRPsMetric = "StrandedEvacuatingActualLRPs"
type EvacuationCleanup struct {
clock clock.Clock
logger lager.Logger
cellID string
exitTimeout time.Duration
bbsClient bbs.InternalClient
executorClient executor.Client
metronClient loggingclient.IngressClient
}
func NewEvacuationCleanup(
logger lager.Logger,
cellID string,
gracefulShutdownInterval time.Duration,
proxyReloadDuration time.Duration,
bbsClient bbs.InternalClient,
executorClient executor.Client,
clock clock.Clock,
metronClient loggingclient.IngressClient,
) *EvacuationCleanup {
return &EvacuationCleanup{
logger: logger,
cellID: cellID,
exitTimeout: gracefulShutdownInterval + proxyReloadDuration + exitTimeoutOffset,
bbsClient: bbsClient,
executorClient: executorClient,
clock: clock,
metronClient: metronClient,
}
}
func (e *EvacuationCleanup) Run(signals <-chan os.Signal, ready chan<- struct{}) error {
logger := e.logger.Session("evacuation-cleanup")
logger.Info("started")
defer logger.Info("complete")
close(ready)
select {
case signal := <-signals:
logger.Info("signalled", lager.Data{"signal": signal})
}
traceID := "" // evacutaion cleanup is not originated through API
actualLRPs, err := e.bbsClient.ActualLRPs(logger, traceID, models.ActualLRPFilter{CellID: e.cellID})
if err != nil {
logger.Error("failed-fetching-actual-lrp-groups", err)
return err
}
strandedEvacuationCount := 0
for _, actualLRP := range actualLRPs {
if actualLRP.GetPresence() != models.ActualLRP_Evacuating {
continue
}
strandedEvacuationCount++
err = e.bbsClient.RemoveEvacuatingActualLRP(logger, traceID, &actualLRP.ActualLRPKey, &actualLRP.ActualLRPInstanceKey)
if err != nil {
logger.Error("failed-removing-evacuating-actual-lrp", err, lager.Data{"lrp-key": actualLRP.ActualLRPKey})
}
}
err = e.metronClient.SendMetric(strandedEvacuatingActualLRPsMetric, strandedEvacuationCount)
if err != nil {
logger.Error("failed-sending-stranded-evacuating-lrp-metric", err, lager.Data{"count": strandedEvacuationCount})
}
logger.Info("finished-evacuating", lager.Data{"stranded-evacuating-actual-lrps": strandedEvacuationCount})
logger.Info("deleting-all-containers")
exitTimer := e.clock.NewTimer(e.exitTimeout)
checkRunningContainersTimer := e.clock.NewTicker(1 * time.Second)
containersSignalled := make(chan struct{})
containersDeleted := make(chan struct{})
go e.deleteRunningContainers(logger, traceID, containersSignalled)
go e.checkRunningContainers(logger, checkRunningContainersTimer.C(), containersSignalled, containersDeleted)
select {
case <-exitTimer.C():
logger.Info("failed-to-cleanup-all-containers")
return errors.New("failed-to-cleanup-all-containers")
case <-containersDeleted:
logger.Info("deleted-containers-successfully")
return nil
}
}
func (e *EvacuationCleanup) checkRunningContainers(
logger lager.Logger,
ticker <-chan time.Time,
containersSignalled <-chan struct{},
containersDeleted chan<- struct{},
) {
hasRunningContainers := func() bool {
containers, err := e.executorClient.ListContainers(logger)
if err != nil {
logger.Error("failed-listing-containers", err)
// assume no container is running if we can't list them
return false
}
if len(containers) > 0 {
return true
} else {
return false
}
}
defer close(containersDeleted)
// wait for all containers to be signalled, this only makes the tests easier
// to write since they depend on the signalling and checking to happen
// sequentially, but isn't necessary for the operation of the cleanup
<-containersSignalled
for hasRunningContainers() {
logger.Info("waiting-for-containers-to-delete")
<-ticker
}
}
func (e *EvacuationCleanup) deleteRunningContainers(logger lager.Logger, traceID string, containersSignalled chan<- struct{}) {
defer close(containersSignalled)
containers, err := e.executorClient.ListContainers(logger)
if err != nil {
logger.Error("failed-listing-containers", err)
return
}
logger.Info("sending-signal-to-containers")
var wg sync.WaitGroup
for _, container := range containers {
sourceName, tags := container.RunInfo.LogConfig.GetSourceNameAndTagsForLogging()
e.metronClient.SendAppLog(fmt.Sprintf("Cell %s reached evacuation timeout for instance %s", e.cellID, container.Guid), sourceName, tags)
wg.Add(1)
go func(logger lager.Logger, traceID string, containerGuid string) {
defer wg.Done()
err := e.executorClient.DeleteContainer(logger, traceID, containerGuid)
if err != nil {
logger.Error("failed-to-delete-container", err, lager.Data{"container-guid": containerGuid})
}
}(logger, traceID, container.Guid)
}
logger.Info("sent-signal-to-containers")
wg.Wait()
}