-
Notifications
You must be signed in to change notification settings - Fork 315
/
batch_resource.go
384 lines (322 loc) · 13 KB
/
batch_resource.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
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
/*
Copyright 2022 The Koordinator 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 batchresource
import (
"fmt"
"math"
utilerrors "k8s.io/apimachinery/pkg/util/errors"
"k8s.io/klog/v2"
"k8s.io/utils/pointer"
apiext "github.com/koordinator-sh/koordinator/apis/extension"
"github.com/koordinator-sh/koordinator/pkg/koordlet/resourceexecutor"
"github.com/koordinator-sh/koordinator/pkg/koordlet/runtimehooks/hooks"
"github.com/koordinator-sh/koordinator/pkg/koordlet/runtimehooks/protocol"
"github.com/koordinator-sh/koordinator/pkg/koordlet/runtimehooks/reconciler"
"github.com/koordinator-sh/koordinator/pkg/koordlet/runtimehooks/rule"
"github.com/koordinator-sh/koordinator/pkg/koordlet/statesinformer"
sysutil "github.com/koordinator-sh/koordinator/pkg/koordlet/util/system"
rmconfig "github.com/koordinator-sh/koordinator/pkg/runtimeproxy/config"
"github.com/koordinator-sh/koordinator/pkg/util"
)
const (
name = "BatchResource"
description = "set fundamental cgroups value for batch pod"
ruleNameForNodeSLO = name + " (nodeSLO)"
ruleNameForNodeMeta = name + " (nodeMeta)"
)
type plugin struct {
rule *Rule
executor resourceexecutor.ResourceUpdateExecutor
}
var podQOSConditions = []string{string(apiext.QoSBE)}
func (p *plugin) Register(op hooks.Options) {
klog.V(5).Infof("register hook %v", name)
rule.Register(ruleNameForNodeSLO, description,
rule.WithParseFunc(statesinformer.RegisterTypeNodeSLOSpec, p.parseRuleForNodeSLO),
rule.WithUpdateCallback(p.ruleUpdateCbForNodeSLO))
rule.Register(ruleNameForNodeMeta, description,
rule.WithParseFunc(statesinformer.RegisterTypeNodeMetadata, p.parseRuleForNodeMeta),
rule.WithUpdateCallback(p.ruleUpdateCbForNodeMeta))
hooks.Register(rmconfig.PreRunPodSandbox, name, description+" (pod)", p.SetPodResources)
hooks.Register(rmconfig.PreCreateContainer, name, description+" (container)", p.SetContainerResources)
hooks.Register(rmconfig.PreUpdateContainerResources, name, description+" (container)", p.SetContainerResources)
reconciler.RegisterCgroupReconciler(reconciler.PodLevel, sysutil.CPUShares, description+" (pod cpu shares)",
p.SetPodCPUShares, reconciler.PodQOSFilter(), podQOSConditions...)
reconciler.RegisterCgroupReconciler(reconciler.PodLevel, sysutil.CPUCFSQuota, description+" (pod cfs quota)",
p.SetPodCFSQuota, reconciler.PodQOSFilter(), podQOSConditions...)
reconciler.RegisterCgroupReconciler(reconciler.PodLevel, sysutil.MemoryLimit, description+" (pod memory limit)",
p.SetPodMemoryLimit, reconciler.PodQOSFilter(), podQOSConditions...)
reconciler.RegisterCgroupReconciler(reconciler.ContainerLevel, sysutil.CPUShares, description+" (container cpu shares)",
p.SetContainerCPUShares, reconciler.PodQOSFilter(), podQOSConditions...)
reconciler.RegisterCgroupReconciler(reconciler.ContainerLevel, sysutil.CPUCFSQuota, description+" (container cfs quota)",
p.SetContainerCFSQuota, reconciler.PodQOSFilter(), podQOSConditions...)
reconciler.RegisterCgroupReconciler(reconciler.ContainerLevel, sysutil.MemoryLimit, description+" (container memory limit)",
p.SetContainerMemoryLimit, reconciler.PodQOSFilter(), podQOSConditions...)
p.executor = op.Executor
}
var singleton *plugin
func Object() *plugin {
if singleton == nil {
singleton = newPlugin()
}
return singleton
}
func newPlugin() *plugin {
return &plugin{
rule: newRule(),
}
}
func (p *plugin) SetPodResources(proto protocol.HooksProtocol) error {
podCtx := proto.(*protocol.PodContext)
if podCtx == nil {
return fmt.Errorf("pod protocol is nil for plugin %v", name)
}
err := p.SetPodCPUShares(proto)
if err != nil {
klog.V(5).Infof("failed to set pod cpu shares in plugin %s, pod %s/%s, err: %v",
name, podCtx.Request.PodMeta.Namespace, podCtx.Request.PodMeta.Name, err)
}
err1 := p.SetPodCFSQuota(proto)
if err1 != nil {
klog.V(5).Infof("failed to set pod cfs quota in plugin %s, pod %s/%s, err: %v",
name, podCtx.Request.PodMeta.Namespace, podCtx.Request.PodMeta.Name, err1)
}
err2 := p.SetPodMemoryLimit(proto)
if err2 != nil {
klog.V(5).Infof("failed to set pod memory limit in plugin %s, pod %s/%s, err: %v",
name, podCtx.Request.PodMeta.Namespace, podCtx.Request.PodMeta.Name, err2)
}
return utilerrors.NewAggregate([]error{err, err1, err2})
}
func (p *plugin) SetPodCPUShares(proto protocol.HooksProtocol) error {
podCtx := proto.(*protocol.PodContext)
if podCtx == nil {
return fmt.Errorf("pod protocol is nil for plugin %v", name)
}
if !isPodQoSBEByAttr(podCtx.Request.Labels, podCtx.Request.Annotations) {
return nil
}
extendedResourceSpec := podCtx.Request.ExtendedResources
// if the extendedResourceSpec is empty, do nothing and keep the original cgroup configs
if extendedResourceSpec == nil {
return nil
}
milliCPURequest := int64(0)
// TODO: count init container and pod overhead
for _, c := range extendedResourceSpec.Containers {
if c.Requests == nil {
continue
}
containerRequest := util.GetBatchMilliCPUFromResourceList(c.Requests)
if containerRequest <= 0 {
continue
}
milliCPURequest += containerRequest
}
cpuShares := sysutil.MilliCPUToShares(milliCPURequest)
podCtx.Response.Resources.CPUShares = pointer.Int64(cpuShares)
return nil
}
func (p *plugin) SetPodCFSQuota(proto protocol.HooksProtocol) error {
podCtx := proto.(*protocol.PodContext)
if podCtx == nil {
return fmt.Errorf("pod protocol is nil for plugin %v", name)
}
if !isPodQoSBEByAttr(podCtx.Request.Labels, podCtx.Request.Annotations) {
return nil
}
extendedResourceSpec := podCtx.Request.ExtendedResources
// if the extendedResourceSpec is empty, do nothing and keep the original cgroup configs
if extendedResourceSpec == nil {
return nil
}
isCFSQuotaEnabled, scaleRatio := p.rule.GetCFSQuotaScaleRatio()
// if cfs quota is disabled, set as -1
if !isCFSQuotaEnabled {
podCtx.Response.Resources.CFSQuota = pointer.Int64(-1)
klog.V(5).Infof("try to unset pod-level cfs quota since it is disabled in plugin %v rule, pod %s/%s",
name, podCtx.Request.PodMeta.Namespace, podCtx.Request.PodMeta.Name)
return nil
}
milliCPULimit := int64(0)
// TODO: count init container and pod overhead
for _, c := range extendedResourceSpec.Containers {
if c.Limits == nil {
milliCPULimit = -1
break
}
containerLimit := util.GetBatchMilliCPUFromResourceList(c.Limits)
if containerLimit <= 0 { // pod unlimited once a container is unlimited
milliCPULimit = -1
break
}
milliCPULimit += containerLimit
}
cfsQuota := sysutil.MilliCPUToQuota(milliCPULimit)
if cfsQuota > 0 && scaleRatio > 1.0 { // no support ratio in (0, 1) yet
originalCFSQuota := cfsQuota
cfsQuota = int64(math.Ceil(float64(originalCFSQuota) / scaleRatio))
klog.V(6).Infof("plugin %s adjusts BE pod %s/%s cfs quota from %d to %d",
name, podCtx.Request.PodMeta.Namespace, podCtx.Request.PodMeta.Name, originalCFSQuota, cfsQuota)
}
podCtx.Response.Resources.CFSQuota = pointer.Int64(cfsQuota)
return nil
}
func (p *plugin) SetPodMemoryLimit(proto protocol.HooksProtocol) error {
podCtx := proto.(*protocol.PodContext)
if podCtx == nil {
return fmt.Errorf("pod protocol is nil for plugin %v", name)
}
if !isPodQoSBEByAttr(podCtx.Request.Labels, podCtx.Request.Annotations) {
return nil
}
extendedResourceSpec := podCtx.Request.ExtendedResources
// if the extendedResourceSpec is empty, do nothing and keep the original cgroup configs
if extendedResourceSpec == nil {
return nil
}
memoryLimit := int64(0)
// TODO: count init container and pod overhead
for _, c := range extendedResourceSpec.Containers {
if c.Limits == nil {
memoryLimit = -1
break
}
containerLimit := util.GetBatchMemoryFromResourceList(c.Limits)
if containerLimit <= 0 { // pod unlimited once a container is unlimited
memoryLimit = -1
break
}
memoryLimit += containerLimit
}
podCtx.Response.Resources.MemoryLimit = pointer.Int64(memoryLimit)
return nil
}
func (p *plugin) SetContainerResources(proto protocol.HooksProtocol) error {
containerCtx := proto.(*protocol.ContainerContext)
if containerCtx == nil {
return fmt.Errorf("container protocol is nil for plugin %v", name)
}
err := p.SetContainerCPUShares(proto)
if err != nil {
klog.V(5).Infof("failed to set container cpu shares in plugin %s, container %s/%s/%s, err: %v",
name, containerCtx.Request.PodMeta.Namespace, containerCtx.Request.PodMeta.Name,
containerCtx.Request.ContainerMeta.Name, err)
}
err1 := p.SetContainerCFSQuota(proto)
if err1 != nil {
klog.V(5).Infof("failed to set container cfs quota in plugin %s, container %s/%s/%s, err: %v",
name, containerCtx.Request.PodMeta.Namespace, containerCtx.Request.PodMeta.Name,
containerCtx.Request.ContainerMeta.Name, err1)
}
err2 := p.SetContainerMemoryLimit(proto)
if err2 != nil {
klog.V(5).Infof("failed to set container memory limit in plugin %s, container %s/%s/%s, err: %v",
name, containerCtx.Request.PodMeta.Namespace, containerCtx.Request.PodMeta.Name,
containerCtx.Request.ContainerMeta.Name, err2)
}
return utilerrors.NewAggregate([]error{err, err1, err2})
}
func (p *plugin) SetContainerCPUShares(proto protocol.HooksProtocol) error {
containerCtx := proto.(*protocol.ContainerContext)
if containerCtx == nil {
return fmt.Errorf("container protocol is nil for plugin %v", name)
}
if !isPodQoSBEByAttr(containerCtx.Request.PodLabels, containerCtx.Request.PodAnnotations) {
return nil
}
containerSpec := containerCtx.Request.ExtendedResources
// if the extendedResourceSpec is empty, do nothing and keep the original cgroup configs
if containerSpec == nil {
return nil
}
milliCPURequest := int64(0)
if containerSpec.Requests != nil {
containerRequest := util.GetBatchMilliCPUFromResourceList(containerSpec.Requests)
if containerRequest > 0 {
milliCPURequest = containerRequest
}
}
cpuShares := sysutil.MilliCPUToShares(milliCPURequest)
containerCtx.Response.Resources.CPUShares = pointer.Int64(cpuShares)
return nil
}
func (p *plugin) SetContainerCFSQuota(proto protocol.HooksProtocol) error {
containerCtx := proto.(*protocol.ContainerContext)
if containerCtx == nil {
return fmt.Errorf("container protocol is nil for plugin %v", name)
}
if !isPodQoSBEByAttr(containerCtx.Request.PodLabels, containerCtx.Request.PodAnnotations) {
return nil
}
containerSpec := containerCtx.Request.ExtendedResources
// if the extendedResourceSpec is empty, do nothing and keep the original cgroup configs
if containerSpec == nil {
return nil
}
isCFSQuotaEnabled, scaleRatio := p.rule.GetCFSQuotaScaleRatio()
// if cfs quota is disabled, set as -1
if !isCFSQuotaEnabled {
containerCtx.Response.Resources.CFSQuota = pointer.Int64(-1)
klog.V(5).Infof("try to unset container-level cfs quota since it is disabled in plugin %v rule, container %s/%s/%s",
name, containerCtx.Request.PodMeta.Namespace, containerCtx.Request.PodMeta.Name,
containerCtx.Request.ContainerMeta.Name)
return nil
}
milliCPULimit := int64(0)
if containerSpec.Limits != nil {
containerLimit := util.GetBatchMilliCPUFromResourceList(containerSpec.Limits)
if containerLimit > 0 {
milliCPULimit = containerLimit
}
}
cfsQuota := sysutil.MilliCPUToQuota(milliCPULimit)
if cfsQuota > 0 && scaleRatio > 1.0 { // no support ratio in (0, 1) yet
originalCFSQuota := cfsQuota
cfsQuota = int64(math.Ceil(float64(originalCFSQuota) / scaleRatio))
klog.V(6).Infof("plugin %s adjusts BE container %s/%s/%s cfs quota from %d to %d",
name, containerCtx.Request.PodMeta.Namespace, containerCtx.Request.PodMeta.Name,
containerCtx.Request.ContainerMeta.Name, originalCFSQuota, cfsQuota)
}
containerCtx.Response.Resources.CFSQuota = pointer.Int64(cfsQuota)
return nil
}
func (p *plugin) SetContainerMemoryLimit(proto protocol.HooksProtocol) error {
containerCtx := proto.(*protocol.ContainerContext)
if containerCtx == nil {
return fmt.Errorf("container protocol is nil for plugin %v", name)
}
if !isPodQoSBEByAttr(containerCtx.Request.PodLabels, containerCtx.Request.PodAnnotations) {
return nil
}
containerSpec := containerCtx.Request.ExtendedResources
// if the extendedResourceSpec is empty, do nothing and keep the original cgroup configs
if containerSpec == nil {
return nil
}
memoryLimit := int64(0)
if containerSpec.Limits != nil {
containerLimit := util.GetBatchMemoryFromResourceList(containerSpec.Limits)
if containerLimit > 0 {
memoryLimit = containerLimit
}
}
if memoryLimit <= 0 {
memoryLimit = -1
}
containerCtx.Response.Resources.MemoryLimit = pointer.Int64(memoryLimit)
return nil
}
func isPodQoSBEByAttr(labels map[string]string, annotations map[string]string) bool {
return apiext.GetQoSClassByAttrs(labels, annotations) == apiext.QoSBE
}