model

package
v2.0.1 Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Apr 20, 2026 License: AGPL-3.0 Imports: 36 Imported by: 0

Documentation

Index

Constants

View Source 
const (
	LabelNameProfileType        = "__profile_type__"
	LabelNameServiceNamePrivate = "__service_name__"
	LabelNameDelta              = "__delta__"
	LabelNameOTEL               = "__otel__"
	LabelNameProfileName        = pmodel.MetricNameLabel
	LabelNamePeriodType         = "__period_type__"
	LabelNamePeriodUnit         = "__period_unit__"
	LabelNameSessionID          = "__session_id__"
	LabelNameType               = "__type__"
	LabelNameUnit               = "__unit__"

	LabelNameServiceGitRef     = "service_git_ref"
	LabelNameServiceName       = "service_name"
	LabelNameServiceRepository = "service_repository"
	LabelNameServiceRootPath   = "service_root_path"

	LabelNameOrder     = "__order__"
	LabelOrderEnforced = "enforced"

	LabelNamePyroscopeSpy = "pyroscope_spy"

	AttrProcessExecutableName = semconv.ProcessExecutableNameKey

	AttrServiceName         = semconv.ServiceNameKey
	AttrServiceNameFallback = "unknown_service"

	ProfileNameOffCpu = "off_cpu" // todo better name?
)
View Source 
const OtherFunctionName = FunctionName(truncatedNodeName)
View Source 
const OtherLocationRef = LocationRefName(-1)
View Source 
const (
	RuleIDLabel = "profiles_rule_id"
)

Variables

This section is empty.

Functions

func CompareLabelPairs 

func CompareLabelPairs(a, b []*typesv1.LabelPair) int

CompareLabelPairs compares the two label sets. The result will be 0 if a==b, <0 if a < b, and >0 if a > b.

func CompareLabelPairs2 

func CompareLabelPairs2(a, b *typesv1.LabelPair) int

func CompareLabels 

func CompareLabels(a, b *typesv1.Labels) int

func CompareProfile 

func CompareProfile(a, b *ingestv1.Profile) int64

CompareProfile compares the two profiles.

func ExportDiffToFlamebearer 

func ExportDiffToFlamebearer(fg *querierv1.FlameGraphDiff, profileType *typesv1.ProfileType) *flamebearer.FlamebearerProfile

func ExportToFlamebearer 

func ExportToFlamebearer(fg *querierv1.FlameGraph, profileType *typesv1.ProfileType) *flamebearer.FlamebearerProfile

ExportToFlamebearer exports the flamegraph to a Flamebearer struct.

func GetSafeTimeRange 

func GetSafeTimeRange(now time.Time, req any) model.Interval

func GetTimeRange 

func GetTimeRange(req TimeRangeRequest) (model.Interval, bool)

GetTimeRange returns the time interval and true if the request has an interval, otherwise ok is false.

func IsLabelAllowedForIngestion 

func IsLabelAllowedForIngestion(name string) bool

func LabelPairsString 

func LabelPairsString(lbs []*typesv1.LabelPair) string

LabelPairsString returns a string representation of the label pairs.

func NewFlameGraph 

func NewFlameGraph(t *FunctionNameTree, maxNodes int64) *querierv1.FlameGraph

func NewFlamegraphDiff 

func NewFlamegraphDiff(left, right *FunctionNameTree, maxNodes int64) (*querierv1.FlameGraphDiff, error)

NewFlamegraphDiff generates a FlameGraphDiff from 2 trees. It also prunes the final tree based on the maxNodes parameter Notice that the resulting FlameGraph can't be used interchangeably with a 'single' Flamegraph Due to many differences: * Nodes * It's structure is different 

i+0 = x offset, left  tree
i+1 = total   , left  tree
i+2 = self    , left  tree
i+3 = x offset, right tree
i+4 = total   , right tree
i+5 = self    , right tree
i+6 = index in the names array

func NewFlamegraphDiffFromBytes 

func NewFlamegraphDiffFromBytes(left, right []byte, maxNodes int64) (*querierv1.FlameGraphDiff, error)

func NewMergeIterator 

func NewMergeIterator[P Profile](max P, deduplicate bool, iters ...iter.Iterator[P]) iter.Iterator[P]

NewMergeIterator returns an iterator that k-way merges the given iterators. The given iterators must be sorted by timestamp and labels themselves. Optionally, the iterator can deduplicate profiles with the same timestamp and labels.

func NewTimeRangedIterator 

func NewTimeRangedIterator[T Timestamp](it iter.Iterator[T], min, max model.Time) iter.Iterator[T]

func ParseMetricSelector 

func ParseMetricSelector(input string) ([]*labels.Matcher, error)

ParseMetricSelector parses the provided textual metric selector into a list of label matchers.

func ParseProfileTypeSelector 

func ParseProfileTypeSelector(id string) (*typesv1.ProfileType, error)

ParseProfileTypeSelector parses the profile selector string.

func SelectorFromProfileType 

func SelectorFromProfileType(profileType *typesv1.ProfileType) *labels.Matcher

SelectorFromProfileType builds a *label.Matcher from an profile type struct

func SetTimeRange 

func SetTimeRange(r interface{}, startTime, endTime model.Time) bool

func SymbolsPartitionForProfile 

func SymbolsPartitionForProfile(ls Labels, partitionLabel string, p *profilev1.Profile) uint64

func TreeFromBackendProfile 

func TreeFromBackendProfile(profile *profilev1.Profile, maxNodes int64) ([]byte, error)

TreeFromBackendProfile is a wrapper...

func TreeFromBackendProfileSampleType 

func TreeFromBackendProfileSampleType(profile *profilev1.Profile, maxNodes int64, sampleType int) ([]byte, error)

TreeFromBackendProfileSampleType converts a pprof profile to a tree format with maxNodes limit

func ValidateMetricName 

func ValidateMetricName(name string) error

Types

type FlameGraphMerger 

type FlameGraphMerger struct {
	// contains filtered or unexported fields
}

func NewFlameGraphMerger 

func NewFlameGraphMerger() *FlameGraphMerger

func (*FlameGraphMerger) FlameGraph 

func (m *FlameGraphMerger) FlameGraph(maxNodes int64) *querierv1.FlameGraph

func (*FlameGraphMerger) MergeFlameGraph 

func (m *FlameGraphMerger) MergeFlameGraph(src *querierv1.FlameGraph)

MergeFlameGraph adds the flame graph stack traces to the resulting flame graph. The call is thread-safe, but the resulting flame graph or tree should be only accessed after all the samples are merged.

func (*FlameGraphMerger) MergeTreeBytes 

func (m *FlameGraphMerger) MergeTreeBytes(src []byte) error

func (*FlameGraphMerger) Tree 

func (m *FlameGraphMerger) Tree() *FunctionNameTree

type FunctionName 

type FunctionName string

type FunctionNameI 

type FunctionNameI struct {
}

func (FunctionNameI) IsLocationTree 

func (FunctionNameI) IsLocationTree() bool

type FunctionNameTree 

type FunctionNameTree = Tree[FunctionName, FunctionNameI]

type LabelMerger 

type LabelMerger struct {
	// contains filtered or unexported fields
}

func NewLabelMerger 

func NewLabelMerger() *LabelMerger

func (*LabelMerger) HasNames 

func (m *LabelMerger) HasNames() bool

func (*LabelMerger) HasSeries 

func (m *LabelMerger) HasSeries() bool

func (*LabelMerger) HasValues 

func (m *LabelMerger) HasValues() bool

func (*LabelMerger) LabelNames 

func (m *LabelMerger) LabelNames() []string

func (*LabelMerger) LabelValues 

func (m *LabelMerger) LabelValues() []string

func (*LabelMerger) Labels 

func (m *LabelMerger) Labels() []*typesv1.Labels

func (*LabelMerger) MergeLabelNames 

func (m *LabelMerger) MergeLabelNames(names []string)

func (*LabelMerger) MergeLabelValues 

func (m *LabelMerger) MergeLabelValues(values []string)

func (*LabelMerger) MergeLabels 

func (m *LabelMerger) MergeLabels(ls []Labels)

func (*LabelMerger) MergeSeries 

func (m *LabelMerger) MergeSeries(series []*typesv1.Labels)

type Labels 

type Labels []*typesv1.LabelPair

Labels is a sorted set of labels. Order has to be guaranteed upon instantiation.

func EmptyLabels 

func EmptyLabels() Labels

EmptyLabels returns n empty Labels value, for convenience.

func IntersectAll 

func IntersectAll(labelSets []Labels) Labels

IntersectAll returns only the labels that are present in all label sets with the same value. Used for merging exemplars with dynamic labels. Reuses the existing Intersect method by iteratively intersecting all label sets.

func LabelsFromMap 

func LabelsFromMap(m map[string]string) Labels

LabelsFromMap returns new sorted Labels from the given map.

func LabelsFromStrings 

func LabelsFromStrings(ss ...string) Labels

LabelsFromStrings creates new labels from pairs of strings.

func (Labels) BytesWithLabels 

func (ls Labels) BytesWithLabels(buf []byte, names ...string) []byte

BytesWithLabels is just as Bytes(), but only for labels matching names. It uses an byte invalid character as a separator and so should not be used for printing.

func (Labels) Clone 

func (ls Labels) Clone() Labels

func (Labels) Delete 

func (ls Labels) Delete(name string) Labels

Delete removes the first label encountered with the name given in place.

func (Labels) Get 

func (ls Labels) Get(name string) string

Get returns the value for the label with the given name. Returns an empty string if the label doesn't exist.

func (Labels) GetLabel 

func (ls Labels) GetLabel(name string) (*typesv1.LabelPair, bool)

GetLabel returns the label with the given name.

func (Labels) Hash 

func (ls Labels) Hash() uint64

Hash returns a hash value for the label set.

func (Labels) InsertSorted 

func (ls Labels) InsertSorted(name, value string) Labels

InsertSorted adds the given label to the set of labels. It assumes the labels are sorted lexicographically.

func (Labels) Intersect 

func (ls Labels) Intersect(labels Labels) Labels

func (Labels) Len 

func (ls Labels) Len() int

func (Labels) Less 

func (ls Labels) Less(i, j int) bool

func (Labels) Range 

func (ls Labels) Range(f func(l *typesv1.LabelPair))

Range calls f on each label.

func (Labels) Subtract 

func (ls Labels) Subtract(labels Labels) Labels

func (Labels) Swap 

func (ls Labels) Swap(i, j int)

func (Labels) ToPrometheusLabels 

func (ls Labels) ToPrometheusLabels() labels.Labels

func (Labels) Unique 

func (ls Labels) Unique() Labels

Unique returns a set labels with unique keys. Labels expected to be sorted: underlying capacity is reused and the original order is preserved: the first key takes precedence over duplicates. Method receiver should not be used after the call.

func (Labels) WithLabels 

func (ls Labels) WithLabels(names ...string) Labels

WithLabels returns a subset of Labels that match with the provided label names.

func (Labels) WithoutLabels 

func (ls Labels) WithoutLabels(names ...string) Labels

WithoutLabels returns a new Labels with the specified label names removed.

func (Labels) WithoutPrivateLabels 

func (ls Labels) WithoutPrivateLabels() Labels

type LabelsBuilder 

type LabelsBuilder struct {
	// contains filtered or unexported fields
}

LabelsBuilder allows modifying Labels.

func NewLabelsBuilder 

func NewLabelsBuilder(base Labels) *LabelsBuilder

NewLabelsBuilder returns a new LabelsBuilder.

func (*LabelsBuilder) Del 

func (b *LabelsBuilder) Del(ns ...string) *LabelsBuilder

Del deletes the label of the given name.

func (*LabelsBuilder) Get 

func (b *LabelsBuilder) Get(n string) string

func (*LabelsBuilder) Labels 

func (b *LabelsBuilder) Labels() Labels

Labels returns the labels from the builder. If no modifications were made, the original labels are returned.

func (*LabelsBuilder) LabelsUnsorted 

func (b *LabelsBuilder) LabelsUnsorted() Labels

LabelsUnsorted returns the labels from the builder. If no modifications were made, the original labels are returned.

The order is not deterministic.

func (*LabelsBuilder) Range 

func (b *LabelsBuilder) Range(f func(l *typesv1.LabelPair))

Range calls f on each label in the Builder.

func (*LabelsBuilder) Reset 

func (b *LabelsBuilder) Reset(base Labels)

Reset clears all current state for the builder.

func (*LabelsBuilder) Set 

func (b *LabelsBuilder) Set(n, v string) *LabelsBuilder

Set the name/value pair as a label.

type LabelsEnforcedOrder 

type LabelsEnforcedOrder []*typesv1.LabelPair

LabelsEnforcedOrder is a sort order of labels, where profile type and service name labels always go first. This is crucial for query performance as labels determine the physical order of the profiling data.

func (LabelsEnforcedOrder) Len 

func (ls LabelsEnforcedOrder) Len() int

func (LabelsEnforcedOrder) Less 

func (ls LabelsEnforcedOrder) Less(i, j int) bool

func (LabelsEnforcedOrder) Swap 

func (ls LabelsEnforcedOrder) Swap(i, j int)

type LocationRefName 

type LocationRefName int

type LocationRefNameI 

type LocationRefNameI struct {
}

func (LocationRefNameI) IsLocationTree 

func (LocationRefNameI) IsLocationTree() bool

type LocationRefNameTree 

type LocationRefNameTree = Tree[LocationRefName, LocationRefNameI]

type MergeIterator 

type MergeIterator[P Profile] struct {
	// contains filtered or unexported fields
}

func (*MergeIterator[P]) At 

func (i *MergeIterator[P]) At() P

func (*MergeIterator[P]) Close 

func (i *MergeIterator[P]) Close() error

func (*MergeIterator[P]) Err 

func (i *MergeIterator[P]) Err() error

func (*MergeIterator[P]) Next 

func (i *MergeIterator[P]) Next() bool

type NodeName 

type NodeName interface {
	~string | ~int
}

type NodeNameI 

type NodeNameI[N ~string | ~int] interface {
	IsLocationTree() bool
	// contains filtered or unexported methods
}

type Profile 

type Profile interface {
	Labels() Labels
	Timestamp
}

type RecordingRule 

type RecordingRule struct {
	Matchers       []*labels.Matcher
	GroupBy        []string
	ExternalLabels labels.Labels
	FunctionName   string
}

func NewRecordingRule 

func NewRecordingRule(rule *settingsv1.RecordingRule) (*RecordingRule, error)

type ServiceVersion 

type ServiceVersion struct {
	Repository string `json:"repository,omitempty"`
	GitRef     string `json:"git_ref,omitempty"`
	BuildID    string `json:"build_id,omitempty"`
	RootPath   string `json:"root_path,omitempty"`
}

func ServiceVersionFromLabels 

func ServiceVersionFromLabels(lbls Labels) (ServiceVersion, bool)

ServiceVersionFromLabels Attempts to extract a service version from the given labels. Returns false if no service version was found.

type SessionID 

type SessionID uint64

func ParseSessionID 

func ParseSessionID(s string) (SessionID, error)

func (SessionID) String 

func (s SessionID) String() string

type SpanSelector 

type SpanSelector map[uint64]struct{}

func NewSpanSelector 

func NewSpanSelector(spans []string) (SpanSelector, error)

type Stack 

type Stack[T any] struct {
	// contains filtered or unexported fields
}

Stack is a stack of values. Pushing and popping values is O(1).

func NewStack 

func NewStack[T any]() *Stack[T]

func (*Stack[T]) Pop 

func (s *Stack[T]) Pop() (result T, ok bool)

Pop removes and returns the top value from the stack.

func (*Stack[T]) Push 

func (s *Stack[T]) Push(v T)

Push adds a value to the top of the stack.

func (*Stack[T]) Reset 

func (s *Stack[T]) Reset()

Reset releases the stack's resources.

func (*Stack[T]) Slice 

func (s *Stack[T]) Slice() []T

Slice returns a slice of the values in the stack. The top value of the stack is at the beginning of the slice.

type StacktraceMerger 

type StacktraceMerger struct {
	// contains filtered or unexported fields
}

func NewStackTraceMerger 

func NewStackTraceMerger() *StacktraceMerger

NewStackTraceMerger merges collections of StacktraceSamples. The result is a byte tree representation of the merged samples.

func (*StacktraceMerger) MergeStackTraces 

func (m *StacktraceMerger) MergeStackTraces(stacks []*ingestv1.StacktraceSample, names []string)

MergeStackTraces adds the stack traces to the resulting tree. The call is thread-safe, but the resulting tree bytes should be only build after all the samples are merged. Note that the function may reuse the capacity of the names slice.

func (*StacktraceMerger) Size 

func (m *StacktraceMerger) Size() int

func (*StacktraceMerger) TreeBytes 

func (m *StacktraceMerger) TreeBytes(maxNodes int64) []byte

type StacktraceNode 

type StacktraceNode struct {
	FirstChild  int32
	NextSibling int32
	Parent      int32
	Location    int32
	Value       int64
	Total       int64
}

type StacktraceTree 

type StacktraceTree struct{ Nodes []StacktraceNode }

func NewStacktraceTree 

func NewStacktraceTree(size int) *StacktraceTree

func (*StacktraceTree) Bytes 

func (t *StacktraceTree) Bytes(dst io.Writer, maxNodes int64, funcs []string)

func (*StacktraceTree) Insert 

func (t *StacktraceTree) Insert(locations []int32, value int64) int32

func (*StacktraceTree) LookupLocations 

func (t *StacktraceTree) LookupLocations(dst []uint64, idx int32) []uint64

func (*StacktraceTree) MinValue 

func (t *StacktraceTree) MinValue(maxNodes int64) int64

MinValue returns the minimum "total" value a node in a tree has to have.

func (*StacktraceTree) Reset 

func (t *StacktraceTree) Reset()

func (*StacktraceTree) Traverse 

func (t *StacktraceTree) Traverse(maxNodes int64, fn StacktraceTreeTraverseFn) error

func (*StacktraceTree) Tree 

func (t *StacktraceTree) Tree(maxNodes int64, names []FunctionName) *FunctionNameTree

type StacktraceTreeTraverseFn 

type StacktraceTreeTraverseFn = func(index int32, children []int32) error

type TimeRangeRequest 

type TimeRangeRequest interface {
	GetStart() int64
	GetEnd() int64
}

TimeRangeRequest is a request that has a time interval.

type TimeRangedIterator 

type TimeRangedIterator[T Timestamp] struct {
	iter.Iterator[T]
	// contains filtered or unexported fields
}

func (*TimeRangedIterator[T]) Next 

func (it *TimeRangedIterator[T]) Next() bool

type Timestamp 

type Timestamp interface {
	Timestamp() model.Time
}

type TraceID 

type TraceID [16]byte

func DecodeTraceID 

func DecodeTraceID(s string) (TraceID, error)

func (TraceID) String 

func (t TraceID) String() string

type Tree 

type Tree[N NodeName, I NodeNameI[N]] struct {
	// contains filtered or unexported fields
}

func MustUnmarshalTree 

func MustUnmarshalTree[N NodeName, I NodeNameI[N]](b []byte) *Tree[N, I]

func TreeFromStacktraceTree 

func TreeFromStacktraceTree[N NodeName, I NodeNameI[N]](t *StacktraceTree, maxNodes int64, lookup func(int32) N) *Tree[N, I]

func UnmarshalTree 

func UnmarshalTree[N NodeName, I NodeNameI[N]](b []byte) (*Tree[N, I], error)

func (*Tree[N, I]) Bytes 

func (t *Tree[N, I]) Bytes(maxNodes int64, keepName func(N) N) []byte

Bytes returns marshaled tree byte representation; the number of nodes is limited to maxNodes. The function modifies the tree: truncated nodes are removed from the tree in place.

func (*Tree[N, I]) Fix 

func (t *Tree[N, I]) Fix()

Fix re-establishes order of nodes and merges duplicates.

func (*Tree[N, I]) FormatNodeNames 

func (t *Tree[N, I]) FormatNodeNames(fn func(N) N)

func (*Tree[N, I]) InsertStack 

func (t *Tree[N, I]) InsertStack(v int64, stack ...N)

func (*Tree[N, I]) IterateStacks 

func (t *Tree[N, I]) IterateStacks(cb func(name N, self int64, stack []N))

func (*Tree[N, I]) MarshalTruncate 

func (t *Tree[N, I]) MarshalTruncate(w io.Writer, maxNodes int64, keepName func(N) N) (err error)

MarshalTruncate writes tree byte representation to the writer provider, the number of nodes is limited to maxNodes. The function modifies the tree: truncated nodes are removed from the tree.

func (*Tree[N, I]) Merge 

func (t *Tree[N, I]) Merge(src *Tree[N, I])

func (*Tree[N, I]) String 

func (t *Tree[N, I]) String() string

func (*Tree[N, I]) Total 

func (t *Tree[N, I]) Total() (v int64)

func (*Tree[N, I]) WriteCollapsed 

func (t *Tree[N, I]) WriteCollapsed(dst io.Writer)

type TreeMergeOption 

type TreeMergeOption[N any] func(o *treeMergeOption[N])

func WithTreeMergeFormatNodeNames 

func WithTreeMergeFormatNodeNames[N any](f func(N) N) TreeMergeOption[N]

type TreeMerger 

type TreeMerger[N NodeName, I NodeNameI[N]] struct {
	// contains filtered or unexported fields
}

func NewTreeMerger 

func NewTreeMerger[N NodeName, I NodeNameI[N]]() *TreeMerger[N, I]

func (*TreeMerger[N, I]) IsEmpty 

func (m *TreeMerger[N, I]) IsEmpty() bool

func (*TreeMerger[N, I]) MergeTree 

func (m *TreeMerger[N, I]) MergeTree(t *Tree[N, I])

func (*TreeMerger[N, I]) MergeTreeBytes 

func (m *TreeMerger[N, I]) MergeTreeBytes(b []byte, opts ...TreeMergeOption[N]) error

func (*TreeMerger[N, I]) Tree 

func (m *TreeMerger[N, I]) Tree() *Tree[N, I]

Source Files

View all Source files

Directories

Path Synopsis
Package timeseries provides types for building and aggregating time series data.
 Package timeseries provides types for building and aggregating time series data.
Package timeseriescompact provides types and utilities for working with compact time series data using attribute table references instead of string labels.
 Package timeseriescompact provides types and utilities for working with compact time series data using attribute table references instead of string labels.

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.