init cgroups

8 months ago · c6eed2192d
6 changed files with 338 additions and 0 deletions
--- a/cgroups/controller.go
+++ b/cgroups/controller.go
@ -0,0 +1,33 @@
+package cgroups
+
+import "fmt"
+
+//go:generate stringer -type=Controller -linecomment
+
+const (
+	ControllerUndefined Controller = iota // undefined
+	ControllerPids                        // pids
+	ControllerCPU                         // cpu
+	ControllerMemory                      // memory
+	ControllerRdma                        // rdma
+	ControllerIo                          // io
+)
+
+type Controller int
+
+func ParseController(in string) (Controller, error) {
+	switch in {
+	case ControllerPids.String():
+		return ControllerPids, nil
+	case ControllerCPU.String():
+		return ControllerCPU, nil
+	case ControllerMemory.String():
+		return ControllerMemory, nil
+	case ControllerRdma.String():
+		return ControllerRdma, nil
+	case ControllerIo.String():
+		return ControllerIo, nil
+	default:
+		return ControllerUndefined, fmt.Errorf("Controller(%s)", in)
+	}
+}
--- a/cgroups/controller_string.go
+++ b/cgroups/controller_string.go
@ -0,0 +1,28 @@
+// Code generated by "stringer -type=Controller -linecomment"; DO NOT EDIT.
+
+package cgroups
+
+import "strconv"
+
+func _() {
+	// An "invalid array index" compiler error signifies that the constant values have changed.
+	// Re-run the stringer command to generate them again.
+	var x [1]struct{}
+	_ = x[ControllerUndefined-0]
+	_ = x[ControllerPids-1]
+	_ = x[ControllerCPU-2]
+	_ = x[ControllerMemory-3]
+	_ = x[ControllerRdma-4]
+	_ = x[ControllerIo-5]
+}
+
+const _Controller_name = "undefinedpidscpumemoryrdmaio"
+
+var _Controller_index = [...]uint8{0, 9, 13, 16, 22, 26, 28}
+
+func (i Controller) String() string {
+	if i < 0 || i >= Controller(len(_Controller_index)-1) {
+		return "Controller(" + strconv.FormatInt(int64(i), 10) + ")"
+	}
+	return _Controller_name[_Controller_index[i]:_Controller_index[i+1]]
+}
--- a/cgroups/cpu.go
+++ b/cgroups/cpu.go
@ -0,0 +1,190 @@
+package cgroups
+
+import (
+	"context"
+	"fmt"
+	"sync"
+
+	cgroups "github.com/containerd/cgroups/v3/cgroup2"
+	"github.com/containerd/cgroups/v3/cgroup2/stats"
+	"go.opentelemetry.io/otel"
+	"go.opentelemetry.io/otel/metric"
+)
+
+// Option supports configuring optional settings for runtime metrics.
+type Option interface {
+	apply(*config)
+}
+
+func newConfig(controllers []Controller, opts ...Option) config {
+	cfg := config{
+		controllers: controllers,
+		provider:    otel.GetMeterProvider(),
+		prefix:      "cgroups.",
+	}
+
+	for _, opt := range opts {
+		opt.apply(&cfg)
+	}
+
+	return cfg
+}
+
+type config struct {
+	controllers []Controller
+	provider    metric.MeterProvider
+	prefix      string
+}
+
+type cgroup struct {
+	meter metric.Meter
+	stats func() (*stats.Metrics, error)
+}
+
+func (c *cgroup) cpu(prefix string) (func(context.Context, metric.Observer, *stats.Metrics), []metric.Observable, error) {
+	var (
+		err           error
+		usageUsec     metric.Int64ObservableGauge
+		userUsec      metric.Int64ObservableGauge
+		systemUsec    metric.Int64ObservableGauge
+		nrPeriods     metric.Int64ObservableGauge
+		nrThrottled   metric.Int64ObservableGauge
+		throttledUsec metric.Int64ObservableGauge
+	)
+
+	if usageUsec, err = c.meter.Int64ObservableGauge(
+		prefix+"usage_usec",
+		metric.WithDescription("Total cpu usage"),
+		metric.WithUnit("microseconds"),
+	); err != nil {
+		return nil, nil, fmt.Errorf("usage_usec:%w", err)
+	}
+
+	if userUsec, err = c.meter.Int64ObservableGauge(
+		prefix+"user_usec",
+		metric.WithDescription("Current cpu usage in user space"),
+		metric.WithUnit("microseconds"),
+	); err != nil {
+		return nil, nil, fmt.Errorf("user_usec:%w", err)
+	}
+
+	if systemUsec, err = c.meter.Int64ObservableGauge(
+		prefix+"system_usec",
+		metric.WithDescription("Current cpu usage in kernel space"),
+		metric.WithUnit("microseconds"),
+	); err != nil {
+		return nil, nil, fmt.Errorf("system_usec:%w", err)
+	}
+
+	if nrPeriods, err = c.meter.Int64ObservableGauge(
+		prefix+"nr_periods",
+		metric.WithDescription("Current cpu number of periods (only if controller is enabled)"),
+	); err != nil {
+		return nil, nil, fmt.Errorf("nr_periods:%w", err)
+	}
+
+	if nrThrottled, err = c.meter.Int64ObservableGauge(
+		prefix+"nr_throttled",
+		metric.WithDescription("Total number of times tasks have been throttled (only if controller is enabled)"),
+	); err != nil {
+		return nil, nil, fmt.Errorf("nr_throttled:%w", err)
+	}
+
+	if throttledUsec, err = c.meter.Int64ObservableGauge(
+		prefix+"throttled_usec",
+		metric.WithDescription("Total time duration for which tasks have been throttled. (only if controller is enabled)"),
+		metric.WithUnit("microseconds"),
+	); err != nil {
+		return nil, nil, fmt.Errorf("throttled_usec:%w", err)
+	}
+
+	return func(ctx context.Context, o metric.Observer, m *stats.Metrics) {
+			o.ObserveInt64(usageUsec, int64(m.CPU.UsageUsec))
+			o.ObserveInt64(userUsec, int64(m.CPU.UserUsec))
+			o.ObserveInt64(systemUsec, int64(m.CPU.SystemUsec))
+			o.ObserveInt64(nrPeriods, int64(m.CPU.NrPeriods))
+			o.ObserveInt64(nrThrottled, int64(m.CPU.NrThrottled))
+			o.ObserveInt64(throttledUsec, int64(m.CPU.ThrottledUsec))
+		}, []metric.Observable{
+			usageUsec, userUsec, systemUsec, nrPeriods, nrThrottled, throttledUsec,
+		}, nil
+}
+
+func (c *cgroup) register(prefix string, ctrls ...Controller) (metric.Registration, error) {
+	observable := make([]metric.Observable, 0, len(ctrls))
+	callbacks := make([]func(context.Context, metric.Observer, *stats.Metrics), 0, len(ctrls))
+	for _, ctrl := range ctrls {
+		switch ctrl {
+		case ControllerCPU:
+			cpu, cpus, err := c.cpu(prefix + "cpu.")
+			if err != nil {
+				return nil, fmt.Errorf("controller cpu:%w", err)
+			}
+
+			observable = append(observable, cpus...)
+			callbacks = append(callbacks, cpu)
+		}
+	}
+
+	var lock sync.Mutex
+
+	lock.Lock()
+	defer lock.Unlock()
+
+	ureg, err := c.meter.RegisterCallback(
+		func(ctx context.Context, obsrv metric.Observer) error {
+			lock.Lock()
+			defer lock.Unlock()
+
+			stats, uErr := c.stats()
+			if uErr != nil {
+				return fmt.Errorf("cpu get stat:%w", uErr)
+			}
+
+			for _, callback := range callbacks {
+				callback(ctx, obsrv, stats)
+			}
+
+			return nil
+		},
+		observable...,
+	)
+
+	if err != nil {
+		return nil, fmt.Errorf("register callback:%w", err)
+	}
+
+	return ureg, nil
+}
+
+func Start(group string, opts ...Option) (metric.Registration, error) {
+	manager, err := cgroups.Load(group)
+	if err != nil {
+		return nil, fmt.Errorf("load:%w", err)
+	}
+
+	controllers, cerr := manager.Controllers()
+	if cerr != nil {
+		return nil, fmt.Errorf("controllers:%w", cerr)
+	}
+
+	ctrls := make([]Controller, 0, len(controllers))
+	for _, controller := range controllers {
+		ctrl, err := ParseController(controller)
+		if err == nil {
+			ctrls = append(ctrls, ctrl)
+		}
+	}
+
+	cfg := newConfig(ctrls, opts...)
+
+	cgr := cgroup{
+		meter: cfg.provider.Meter(
+			"gitoa.ru/go-4devs/otel/cgroups",
+			metric.WithInstrumentationVersion(Version()),
+		),
+		stats: manager.Stat,
+	}
+
+	return cgr.register(cfg.prefix+group+".", cfg.controllers...)
+}
--- a/cgroups/go.mod
+++ b/cgroups/go.mod
@ -0,0 +1,23 @@
+module gitoa.ru/go-4devs/otel/cgroups
+
+go 1.22.2
+
+require (
+	github.com/containerd/cgroups/v3 v3.0.3
+	go.opentelemetry.io/otel v1.25.0
+	go.opentelemetry.io/otel/metric v1.25.0
+)
+
+require (
+	github.com/cilium/ebpf v0.11.0 // indirect
+	github.com/coreos/go-systemd/v22 v22.3.2 // indirect
+	github.com/go-logr/logr v1.4.1 // indirect
+	github.com/go-logr/stdr v1.2.2 // indirect
+	github.com/godbus/dbus/v5 v5.0.4 // indirect
+	github.com/opencontainers/runtime-spec v1.0.2 // indirect
+	github.com/sirupsen/logrus v1.9.0 // indirect
+	go.opentelemetry.io/otel/trace v1.25.0 // indirect
+	golang.org/x/exp v0.0.0-20230224173230-c95f2b4c22f2 // indirect
+	golang.org/x/sys v0.6.0 // indirect
+	google.golang.org/protobuf v1.27.1 // indirect
+)
--- a/cgroups/go.sum
+++ b/cgroups/go.sum
@ -0,0 +1,59 @@
+github.com/cilium/ebpf v0.11.0 h1:V8gS/bTCCjX9uUnkUFUpPsksM8n1lXBAvHcpiFk1X2Y=
+github.com/cilium/ebpf v0.11.0/go.mod h1:WE7CZAnqOL2RouJ4f1uyNhqr2P4CCvXFIqdRDUgWsVs=
+github.com/containerd/cgroups/v3 v3.0.3 h1:S5ByHZ/h9PMe5IOQoN7E+nMc2UcLEM/V48DGDJ9kip0=
+github.com/containerd/cgroups/v3 v3.0.3/go.mod h1:8HBe7V3aWGLFPd/k03swSIsGjZhHI2WzJmticMgVuz0=
+github.com/coreos/go-systemd/v22 v22.3.2 h1:D9/bQk5vlXQFZ6Kwuu6zaiXJ9oTPe68++AzAJc1DzSI=
+github.com/coreos/go-systemd/v22 v22.3.2/go.mod h1:Y58oyj3AT4RCenI/lSvhwexgC+NSVTIJ3seZv2GcEnc=
+github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
+github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
+github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
+github.com/frankban/quicktest v1.14.5 h1:dfYrrRyLtiqT9GyKXgdh+k4inNeTvmGbuSgZ3lx3GhA=
+github.com/frankban/quicktest v1.14.5/go.mod h1:4ptaffx2x8+WTWXmUCuVU6aPUX1/Mz7zb5vbUoiM6w0=
+github.com/go-logr/logr v1.2.2/go.mod h1:jdQByPbusPIv2/zmleS9BjJVeZ6kBagPoEUsqbVz/1A=
+github.com/go-logr/logr v1.4.1 h1:pKouT5E8xu9zeFC39JXRDukb6JFQPXM5p5I91188VAQ=
+github.com/go-logr/logr v1.4.1/go.mod h1:9T104GzyrTigFIr8wt5mBrctHMim0Nb2HLGrmQ40KvY=
+github.com/go-logr/stdr v1.2.2 h1:hSWxHoqTgW2S2qGc0LTAI563KZ5YKYRhT3MFKZMbjag=
+github.com/go-logr/stdr v1.2.2/go.mod h1:mMo/vtBO5dYbehREoey6XUKy/eSumjCCveDpRre4VKE=
+github.com/godbus/dbus/v5 v5.0.4 h1:9349emZab16e7zQvpmsbtjc18ykshndd8y2PG3sgJbA=
+github.com/godbus/dbus/v5 v5.0.4/go.mod h1:xhWf0FNVPg57R7Z0UbKHbJfkEywrmjJnf7w5xrFpKfA=
+github.com/golang/protobuf v1.5.0/go.mod h1:FsONVRAS9T7sI+LIUmWTfcYkHO4aIWwzhcaSAoJOfIk=
+github.com/google/go-cmp v0.5.5/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/gNBxE=
+github.com/google/go-cmp v0.6.0 h1:ofyhxvXcZhMsU5ulbFiLKl/XBFqE1GSq7atu8tAmTRI=
+github.com/google/go-cmp v0.6.0/go.mod h1:17dUlkBOakJ0+DkrSSNjCkIjxS6bF9zb3elmeNGIjoY=
+github.com/kr/pretty v0.3.1 h1:flRD4NNwYAUpkphVc1HcthR4KEIFJ65n8Mw5qdRn3LE=
+github.com/kr/pretty v0.3.1/go.mod h1:hoEshYVHaxMs3cyo3Yncou5ZscifuDolrwPKZanG3xk=
+github.com/kr/text v0.2.0 h1:5Nx0Ya0ZqY2ygV366QzturHI13Jq95ApcVaJBhpS+AY=
+github.com/kr/text v0.2.0/go.mod h1:eLer722TekiGuMkidMxC/pM04lWEeraHUUmBw8l2grE=
+github.com/opencontainers/runtime-spec v1.0.2 h1:UfAcuLBJB9Coz72x1hgl8O5RVzTdNiaglX6v2DM6FI0=
+github.com/opencontainers/runtime-spec v1.0.2/go.mod h1:jwyrGlmzljRJv/Fgzds9SsS/C5hL+LL3ko9hs6T5lQ0=
+github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
+github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
+github.com/rogpeppe/go-internal v1.9.0 h1:73kH8U+JUqXU8lRuOHeVHaa/SZPifC7BkcraZVejAe8=
+github.com/rogpeppe/go-internal v1.9.0/go.mod h1:WtVeX8xhTBvf0smdhujwtBcq4Qrzq/fJaraNFVN+nFs=
+github.com/sirupsen/logrus v1.9.0 h1:trlNQbNUG3OdDrDil03MCb1H2o9nJ1x4/5LYw7byDE0=
+github.com/sirupsen/logrus v1.9.0/go.mod h1:naHLuLoDiP4jHNo9R0sCBMtWGeIprob74mVsIT4qYEQ=
+github.com/stretchr/objx v0.1.0/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+wExME=
+github.com/stretchr/testify v1.7.0/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg=
+github.com/stretchr/testify v1.9.0 h1:HtqpIVDClZ4nwg75+f6Lvsy/wHu+3BoSGCbBAcpTsTg=
+github.com/stretchr/testify v1.9.0/go.mod h1:r2ic/lqez/lEtzL7wO/rwa5dbSLXVDPFyf8C91i36aY=
+go.opentelemetry.io/otel v1.25.0 h1:gldB5FfhRl7OJQbUHt/8s0a7cE8fbsPAtdpRaApKy4k=
+go.opentelemetry.io/otel v1.25.0/go.mod h1:Wa2ds5NOXEMkCmUou1WA7ZBfLTHWIsp034OVD7AO+Vg=
+go.opentelemetry.io/otel/metric v1.25.0 h1:LUKbS7ArpFL/I2jJHdJcqMGxkRdxpPHE0VU/D4NuEwA=
+go.opentelemetry.io/otel/metric v1.25.0/go.mod h1:rkDLUSd2lC5lq2dFNrX9LGAbINP5B7WBkC78RXCpH5s=
+go.opentelemetry.io/otel/trace v1.25.0 h1:tqukZGLwQYRIFtSQM2u2+yfMVTgGVeqRLPUYx1Dq6RM=
+go.opentelemetry.io/otel/trace v1.25.0/go.mod h1:hCCs70XM/ljO+BeQkyFnbK28SBIJ/Emuha+ccrCRT7I=
+go.uber.org/goleak v1.1.12 h1:gZAh5/EyT/HQwlpkCy6wTpqfH9H8Lz8zbm3dZh+OyzA=
+go.uber.org/goleak v1.1.12/go.mod h1:cwTWslyiVhfpKIDGSZEM2HlOvcqm+tG4zioyIeLoqMQ=
+golang.org/x/exp v0.0.0-20230224173230-c95f2b4c22f2 h1:Jvc7gsqn21cJHCmAWx0LiimpP18LZmUxkT5Mp7EZ1mI=
+golang.org/x/exp v0.0.0-20230224173230-c95f2b4c22f2/go.mod h1:CxIveKay+FTh1D0yPZemJVgC/95VzuuOLq5Qi4xnoYc=
+golang.org/x/sys v0.0.0-20220715151400-c0bba94af5f8/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
+golang.org/x/sys v0.6.0 h1:MVltZSvRTcU2ljQOhs94SXPftV6DCNnZViHeQps87pQ=
+golang.org/x/sys v0.6.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
+golang.org/x/xerrors v0.0.0-20191204190536-9bdfabe68543/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
+google.golang.org/protobuf v1.26.0-rc.1/go.mod h1:jlhhOSvTdKEhbULTjvd4ARK9grFBp09yW+WbY/TyQbw=
+google.golang.org/protobuf v1.27.1 h1:SnqbnDw1V7RiZcXPx5MEeqPv2s79L9i7BJUlG/+RurQ=
+google.golang.org/protobuf v1.27.1/go.mod h1:9q0QmTI4eRPtz6boOQmLYwt+qCgq0jsYwAQnmE0givc=
+gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
+gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=
+gopkg.in/yaml.v3 v3.0.1 h1:fxVm/GzAzEWqLHuvctI91KS9hhNmmWOoWu0XTYJS7CA=
+gopkg.in/yaml.v3 v3.0.1/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=
--- a/cgroups/version.go
+++ b/cgroups/version.go
@ -0,0 +1,5 @@
+package cgroups
+
+func Version() string {
+	return "0.0.1"
+}