Documentation ¶
Index ¶
- Variables
- func FindMinMaxTime(s *parser.EvalStmt) (int64, int64)
- func HistogramTestExpression(h *histogram.FloatHistogram) string
- func LoadedStorage(t testutil.T, input string) *teststorage.TestStorage
- func NewOriginContext(ctx context.Context, data map[string]interface{}) context.Context
- func PreprocessExpr(expr parser.Expr, start, end time.Time) parser.Expr
- func RunBuiltinTests(t *testing.T, engine engineQuerier)
- func RunTest(t testutil.T, input string, engine engineQuerier)
- type ActiveQueryTracker
- type Engine
- type EngineOpts
- type Entry
- type ErrQueryCanceled
- type ErrQueryTimeout
- type ErrStorage
- type ErrTooManySamples
- type EvalNodeHelper
- type EvalSeriesHelper
- type FPoint
- type FunctionCall
- type HPoint
- type LazyLoader
- type LazyLoaderOpts
- type Matrix
- type PrometheusQueryOpts
- type Query
- type QueryLogger
- type QueryOpts
- type QueryOrigin
- type QueryTracker
- type Result
- type Sample
- type Scalar
- type Series
- type StorageSeries
- type String
- type Vector
Constants ¶
This section is empty.
Variables ¶
var ( ErrValidationAtModifierDisabled = errors.New("@ modifier is disabled") ErrValidationNegativeOffsetDisabled = errors.New("negative offset is disabled") )
var AtModifierUnsafeFunctions = map[string]struct{}{
"days_in_month": {}, "day_of_month": {}, "day_of_week": {}, "day_of_year": {},
"hour": {}, "minute": {}, "month": {}, "year": {},
"predict_linear": {}, "time": {},
"timestamp": {},
}
AtModifierUnsafeFunctions are the functions whose result can vary if evaluation time is changed when the arguments are step invariant. It also includes functions that use the timestamps of the passed instant vector argument to calculate a result since that can also change with change in eval time.
var FunctionCalls = map[string]FunctionCall{
"abs": funcAbs,
"absent": funcAbsent,
"absent_over_time": funcAbsentOverTime,
"acos": funcAcos,
"acosh": funcAcosh,
"asin": funcAsin,
"asinh": funcAsinh,
"atan": funcAtan,
"atanh": funcAtanh,
"avg_over_time": funcAvgOverTime,
"ceil": funcCeil,
"changes": funcChanges,
"clamp": funcClamp,
"clamp_max": funcClampMax,
"clamp_min": funcClampMin,
"cos": funcCos,
"cosh": funcCosh,
"count_over_time": funcCountOverTime,
"days_in_month": funcDaysInMonth,
"day_of_month": funcDayOfMonth,
"day_of_week": funcDayOfWeek,
"day_of_year": funcDayOfYear,
"deg": funcDeg,
"delta": funcDelta,
"deriv": funcDeriv,
"exp": funcExp,
"floor": funcFloor,
"histogram_count": funcHistogramCount,
"histogram_fraction": funcHistogramFraction,
"histogram_quantile": funcHistogramQuantile,
"histogram_sum": funcHistogramSum,
"histogram_stddev": funcHistogramStdDev,
"histogram_stdvar": funcHistogramStdVar,
"holt_winters": funcHoltWinters,
"hour": funcHour,
"idelta": funcIdelta,
"increase": funcIncrease,
"irate": funcIrate,
"label_replace": funcLabelReplace,
"label_join": funcLabelJoin,
"ln": funcLn,
"log10": funcLog10,
"log2": funcLog2,
"last_over_time": funcLastOverTime,
"mad_over_time": funcMadOverTime,
"max_over_time": funcMaxOverTime,
"min_over_time": funcMinOverTime,
"minute": funcMinute,
"month": funcMonth,
"pi": funcPi,
"predict_linear": funcPredictLinear,
"present_over_time": funcPresentOverTime,
"quantile_over_time": funcQuantileOverTime,
"rad": funcRad,
"rate": funcRate,
"resets": funcResets,
"round": funcRound,
"scalar": funcScalar,
"sgn": funcSgn,
"sin": funcSin,
"sinh": funcSinh,
"sort": funcSort,
"sort_desc": funcSortDesc,
"sort_by_label": funcSortByLabel,
"sort_by_label_desc": funcSortByLabelDesc,
"sqrt": funcSqrt,
"stddev_over_time": funcStddevOverTime,
"stdvar_over_time": funcStdvarOverTime,
"sum_over_time": funcSumOverTime,
"tan": funcTan,
"tanh": funcTanh,
"time": funcTime,
"timestamp": funcTimestamp,
"vector": funcVector,
"year": funcYear,
}
FunctionCalls is a list of all functions supported by PromQL, including their types.
Functions ¶
func FindMinMaxTime ¶ added in v0.49.0
FindMinMaxTime returns the time in milliseconds of the earliest and latest point in time the statement will try to process. This takes into account offsets, @ modifiers, and range selectors. If the statement does not select series, then FindMinMaxTime returns (0, 0).
func HistogramTestExpression ¶ added in v0.47.0
func HistogramTestExpression(h *histogram.FloatHistogram) string
HistogramTestExpression returns TestExpression() for the given histogram or "" if the histogram is nil.
func LoadedStorage ¶ added in v0.47.0
func LoadedStorage(t testutil.T, input string) *teststorage.TestStorage
LoadedStorage returns storage with generated data using the provided load statements. Non-load statements will cause test errors.
func NewOriginContext ¶
NewOriginContext returns a new context with data about the origin attached.
func PreprocessExpr ¶
PreprocessExpr wraps all possible step invariant parts of the given expression with StepInvariantExpr. It also resolves the preprocessors.
func RunBuiltinTests ¶ added in v0.47.0
RunBuiltinTests runs an acceptance test suite against the provided engine.
Types ¶
type ActiveQueryTracker ¶
type ActiveQueryTracker struct {
// contains filtered or unexported fields
}
func NewActiveQueryTracker ¶
func NewActiveQueryTracker(localStoragePath string, maxConcurrent int, logger log.Logger) *ActiveQueryTracker
func (*ActiveQueryTracker) Close ¶ added in v0.46.0
func (tracker *ActiveQueryTracker) Close()
func (ActiveQueryTracker) Delete ¶
func (tracker ActiveQueryTracker) Delete(insertIndex int)
func (ActiveQueryTracker) GetMaxConcurrent ¶
func (tracker ActiveQueryTracker) GetMaxConcurrent() int
type Engine ¶
type Engine struct {
// contains filtered or unexported fields
}
Engine handles the lifetime of queries from beginning to end. It is connected to a querier.
func (*Engine) NewInstantQuery ¶
func (ng *Engine) NewInstantQuery(ctx context.Context, q storage.Queryable, opts QueryOpts, qs string, ts time.Time) (Query, error)
NewInstantQuery returns an evaluation query for the given expression at the given time.
func (*Engine) NewRangeQuery ¶
func (ng *Engine) NewRangeQuery(ctx context.Context, q storage.Queryable, opts QueryOpts, qs string, start, end time.Time, interval time.Duration) (Query, error)
NewRangeQuery returns an evaluation query for the given time range and with the resolution set by the interval.
func (*Engine) SetQueryLogger ¶
func (ng *Engine) SetQueryLogger(l QueryLogger)
SetQueryLogger sets the query logger.
type EngineOpts ¶
type EngineOpts struct { Logger log.Logger Reg prometheus.Registerer MaxSamples int Timeout time.Duration ActiveQueryTracker QueryTracker // LookbackDelta determines the time since the last sample after which a time // series is considered stale. LookbackDelta time.Duration // NoStepSubqueryIntervalFn is the default evaluation interval of // a subquery in milliseconds if no step in range vector was specified `[30m:<step>]`. NoStepSubqueryIntervalFn func(rangeMillis int64) int64 // EnableAtModifier if true enables @ modifier. Disabled otherwise. This // is supposed to be enabled for regular PromQL (as of Prometheus v2.33) // but the option to disable it is still provided here for those using // the Engine outside of Prometheus. EnableAtModifier bool // EnableNegativeOffset if true enables negative (-) offset // values. Disabled otherwise. This is supposed to be enabled for // regular PromQL (as of Prometheus v2.33) but the option to disable it // is still provided here for those using the Engine outside of // Prometheus. EnableNegativeOffset bool // EnablePerStepStats if true allows for per-step stats to be computed on request. Disabled otherwise. EnablePerStepStats bool }
EngineOpts contains configuration options used when creating a new Engine.
type ErrQueryCanceled ¶
type ErrQueryCanceled string
ErrQueryCanceled is returned if a query was canceled during processing.
func (ErrQueryCanceled) Error ¶
func (e ErrQueryCanceled) Error() string
type ErrQueryTimeout ¶
type ErrQueryTimeout string
ErrQueryTimeout is returned if a query timed out during processing.
func (ErrQueryTimeout) Error ¶
func (e ErrQueryTimeout) Error() string
type ErrStorage ¶
type ErrStorage struct{ Err error }
ErrStorage is returned if an error was encountered in the storage layer during query handling.
func (ErrStorage) Error ¶
func (e ErrStorage) Error() string
type ErrTooManySamples ¶
type ErrTooManySamples string
ErrTooManySamples is returned if a query would load more than the maximum allowed samples into memory.
func (ErrTooManySamples) Error ¶
func (e ErrTooManySamples) Error() string
type EvalNodeHelper ¶
type EvalNodeHelper struct { // Evaluation timestamp. Ts int64 // Vector that can be used for output. Out Vector // Caches. // DropMetricName and label_*. Dmn map[uint64]labels.Labels // contains filtered or unexported fields }
EvalNodeHelper stores extra information and caches for evaluating a single node across steps.
func (*EvalNodeHelper) DropMetricName ¶
func (enh *EvalNodeHelper) DropMetricName(l labels.Labels) labels.Labels
DropMetricName is a cached version of DropMetricName.
type EvalSeriesHelper ¶
type EvalSeriesHelper struct {
// contains filtered or unexported fields
}
EvalSeriesHelper stores extra information about a series.
type FPoint ¶ added in v0.44.0
FPoint represents a single float data point for a given timestamp.
func (FPoint) MarshalJSON ¶ added in v0.44.0
MarshalJSON implements json.Marshaler.
JSON marshaling is only needed for the HTTP API. Since FPoint is such a frequently marshaled type, it gets an optimized treatment directly in web/api/v1/api.go. Therefore, this method is unused within Prometheus. It is still provided here as convenience for debugging and for other users of this code. Also note that the different marshaling implementations might lead to slightly different results in terms of formatting and rounding of the timestamp.
type FunctionCall ¶
type FunctionCall func(vals []parser.Value, args parser.Expressions, enh *EvalNodeHelper) (Vector, annotations.Annotations)
FunctionCall is the type of a PromQL function implementation
vals is a list of the evaluated arguments for the function call.
For range vectors it will be a Matrix with one series, instant vectors a Vector, scalars a Vector with one series whose value is the scalar value,and nil for strings.
args are the original arguments to the function, where you can access matrixSelectors, vectorSelectors, and StringLiterals.
enh.Out is a pre-allocated empty vector that you may use to accumulate output before returning it. The vectors in vals should not be returned.a
Range vector functions need only return a vector with the right value, the metric and timestamp are not needed.
Instant vector functions need only return a vector with the right values and metrics, the timestamp are not needed.
Scalar results should be returned as the value of a sample in a Vector.
type HPoint ¶ added in v0.44.0
type HPoint struct { T int64 H *histogram.FloatHistogram }
HPoint represents a single histogram data point for a given timestamp. H must never be nil.
func (HPoint) MarshalJSON ¶ added in v0.44.0
MarshalJSON implements json.Marshaler.
JSON marshaling is only needed for the HTTP API. Since HPoint is such a frequently marshaled type, it gets an optimized treatment directly in web/api/v1/api.go. Therefore, this method is unused within Prometheus. It is still provided here as convenience for debugging and for other users of this code. Also note that the different marshaling implementations might lead to slightly different results in terms of formatting and rounding of the timestamp.
type LazyLoader ¶
type LazyLoader struct { testutil.T SubqueryInterval time.Duration // contains filtered or unexported fields }
LazyLoader lazily loads samples into storage. This is specifically implemented for unit testing of rules.
func NewLazyLoader ¶
func NewLazyLoader(t testutil.T, input string, opts LazyLoaderOpts) (*LazyLoader, error)
NewLazyLoader returns an initialized empty LazyLoader.
func (*LazyLoader) Close ¶
func (ll *LazyLoader) Close()
Close closes resources associated with the LazyLoader.
func (*LazyLoader) Context ¶
func (ll *LazyLoader) Context() context.Context
Context returns the LazyLoader's context.
func (*LazyLoader) QueryEngine ¶
func (ll *LazyLoader) QueryEngine() *Engine
QueryEngine returns the LazyLoader's query engine.
func (*LazyLoader) Queryable ¶
func (ll *LazyLoader) Queryable() storage.Queryable
Queryable allows querying the LazyLoader's data. Note: only the samples till the max timestamp used in `WithSamplesTill` can be queried.
func (*LazyLoader) Storage ¶
func (ll *LazyLoader) Storage() storage.Storage
Storage returns the LazyLoader's storage.
func (*LazyLoader) WithSamplesTill ¶
func (ll *LazyLoader) WithSamplesTill(ts time.Time, fn func(error))
WithSamplesTill loads the samples till given timestamp and executes the given function.
type LazyLoaderOpts ¶
type LazyLoaderOpts struct {
// Both of these must be set to true for regular PromQL (as of
// Prometheus v2.33). They can still be disabled here for legacy and
// other uses.
EnableAtModifier, EnableNegativeOffset bool
}
LazyLoaderOpts are options for the lazy loader.
type Matrix ¶
type Matrix []Series
Matrix is a slice of Series that implements sort.Interface and has a String method.
func (Matrix) ContainsSameLabelset ¶
ContainsSameLabelset checks if a matrix has samples with the same labelset. Such a behavior is semantically undefined. https://github.com/prometheus/prometheus/issues/4562
func (Matrix) TotalSamples ¶
TotalSamples returns the total number of samples in the series within a matrix. Float samples have a weight of 1 in this number, while histogram samples have a higher weight according to their size compared with the size of a float sample. See HPoint.size for details.
type PrometheusQueryOpts ¶ added in v0.46.0
type PrometheusQueryOpts struct {
// contains filtered or unexported fields
}
func (*PrometheusQueryOpts) EnablePerStepStats ¶ added in v0.46.0
func (p *PrometheusQueryOpts) EnablePerStepStats() bool
func (*PrometheusQueryOpts) LookbackDelta ¶ added in v0.46.0
func (p *PrometheusQueryOpts) LookbackDelta() time.Duration
type Query ¶
type Query interface { // Exec processes the query. Can only be called once. Exec(ctx context.Context) *Result // Close recovers memory used by the query result. Close() // Statement returns the parsed statement of the query. Statement() parser.Statement // Stats returns statistics about the lifetime of the query. Stats() *stats.Statistics // Cancel signals that a running query execution should be aborted. Cancel() // String returns the original query string. String() string }
A Query is derived from an a raw query string and can be run against an engine it is associated with.
type QueryLogger ¶
QueryLogger is an interface that can be used to log all the queries logged by the engine.
type QueryOpts ¶
type QueryOrigin ¶
type QueryOrigin struct{}
type QueryTracker ¶
type QueryTracker interface { // GetMaxConcurrent returns maximum number of concurrent queries that are allowed by this tracker. GetMaxConcurrent() int // Insert inserts query into query tracker. This call must block if maximum number of queries is already running. // If Insert doesn't return error then returned integer value should be used in subsequent Delete call. // Insert should return error if context is finished before query can proceed, and integer value returned in this case should be ignored by caller. Insert(ctx context.Context, query string) (int, error) // Delete removes query from activity tracker. InsertIndex is value returned by Insert call. Delete(insertIndex int) }
QueryTracker provides access to two features:
1) Tracking of active query. If PromQL engine crashes while executing any query, such query should be present in the tracker on restart, hence logged. After the logging on restart, the tracker gets emptied.
2) Enforcement of the maximum number of concurrent queries.
type Result ¶
type Result struct { Err error Value parser.Value Warnings annotations.Annotations }
Result holds the resulting value of an execution or an error if any occurred.
func (*Result) Matrix ¶
Matrix returns a Matrix. An error is returned if the result was an error or the result value is not a Matrix.
type Sample ¶
Sample is a single sample belonging to a metric. It represents either a float sample or a histogram sample. If H is nil, it is a float sample. Otherwise, it is a histogram sample.
func (Sample) MarshalJSON ¶
MarshalJSON is mirrored in web/api/v1/api.go with jsoniter because FPoint and HPoint wouldn't be marshaled with jsoniter otherwise.
type Scalar ¶
Scalar is a data point that's explicitly not associated with a metric.
func (Scalar) MarshalJSON ¶
type Series ¶
type Series struct { Metric labels.Labels `json:"metric"` Floats []FPoint `json:"values,omitempty"` Histograms []HPoint `json:"histograms,omitempty"` }
Series is a stream of data points belonging to a metric.
type StorageSeries ¶
type StorageSeries struct {
// contains filtered or unexported fields
}
StorageSeries simulates promql.Series as storage.Series.
func NewStorageSeries ¶
func NewStorageSeries(series Series) *StorageSeries
NewStorageSeries returns a StorageSeries from a Series.
func (*StorageSeries) Iterator ¶
func (ss *StorageSeries) Iterator(it chunkenc.Iterator) chunkenc.Iterator
Iterator returns a new iterator of the data of the series. In case of multiple samples with the same timestamp, it returns the float samples first.
func (*StorageSeries) Labels ¶
func (ss *StorageSeries) Labels() labels.Labels
type Vector ¶
type Vector []Sample
Vector is basically only an alias for []Sample, but the contract is that in a Vector, all Samples have the same timestamp.
func (Vector) ContainsSameLabelset ¶
ContainsSameLabelset checks if a vector has samples with the same labelset Such a behavior is semantically undefined https://github.com/prometheus/prometheus/issues/4562
func (Vector) TotalSamples ¶ added in v0.49.0
TotalSamples returns the total number of samples in the series within a vector. Float samples have a weight of 1 in this number, while histogram samples have a higher weight according to their size compared with the size of a float sample. See HPoint.size for details.