You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
prometheus/rules/helpers.go

222 lines
7.7 KiB

// Copyright 2013 The Prometheus 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 rules
import (
"fmt"
"net/url"
"strings"
clientmodel "github.com/prometheus/client_golang/model"
"github.com/prometheus/prometheus/rules/ast"
"github.com/prometheus/prometheus/storage/metric"
"github.com/prometheus/prometheus/utility"
)
// CreateRecordingRule is a convenience function to create a recording rule.
func CreateRecordingRule(name string, labels clientmodel.LabelSet, expr ast.Node, permanent bool) (*RecordingRule, error) {
if _, ok := expr.(ast.VectorNode); !ok {
return nil, fmt.Errorf("recording rule expression %v does not evaluate to vector type", expr)
}
return &RecordingRule{
name: name,
labels: labels,
vector: expr.(ast.VectorNode),
permanent: permanent,
}, nil
}
// CreateAlertingRule is a convenience function to create a new alerting rule.
func CreateAlertingRule(name string, expr ast.Node, holdDurationStr string, labels clientmodel.LabelSet, summary string, description string) (*AlertingRule, error) {
if _, ok := expr.(ast.VectorNode); !ok {
return nil, fmt.Errorf("alert rule expression %v does not evaluate to vector type", expr)
}
holdDuration, err := utility.StringToDuration(holdDurationStr)
if err != nil {
return nil, err
}
return NewAlertingRule(name, expr.(ast.VectorNode), holdDuration, labels, summary, description), nil
}
// NewScalarLiteral returns a ScalarLiteral with the given value. If sign is "-"
// the value is negated.
func NewScalarLiteral(value clientmodel.SampleValue, sign string) *ast.ScalarLiteral {
if sign == "-" {
value = -value
}
return ast.NewScalarLiteral(value)
}
// NewFunctionCall is a convenience function to create a new AST function-call node.
func NewFunctionCall(name string, args []ast.Node) (ast.Node, error) {
function, err := ast.GetFunction(name)
if err != nil {
return nil, fmt.Errorf("unknown function %q", name)
}
functionCall, err := ast.NewFunctionCall(function, args)
if err != nil {
return nil, fmt.Errorf(err.Error())
}
return functionCall, nil
}
// NewVectorAggregation is a convenience function to create a new AST vector aggregation.
func NewVectorAggregation(aggrTypeStr string, vector ast.Node, groupBy clientmodel.LabelNames, keepExtraLabels bool) (*ast.VectorAggregation, error) {
if _, ok := vector.(ast.VectorNode); !ok {
return nil, fmt.Errorf("operand of %v aggregation must be of vector type", aggrTypeStr)
}
var aggrTypes = map[string]ast.AggrType{
"SUM": ast.Sum,
"MAX": ast.Max,
"MIN": ast.Min,
"AVG": ast.Avg,
"COUNT": ast.Count,
}
aggrType, ok := aggrTypes[aggrTypeStr]
if !ok {
return nil, fmt.Errorf("unknown aggregation type %q", aggrTypeStr)
}
return ast.NewVectorAggregation(aggrType, vector.(ast.VectorNode), groupBy, keepExtraLabels), nil
}
// vectorMatching combines data used to match samples between vectors.
type vectorMatching struct {
matchCardinality ast.VectorMatchCardinality
matchOn clientmodel.LabelNames
includeLabels clientmodel.LabelNames
}
// newVectorMatching is a convenience function to create a new vectorMatching.
func newVectorMatching(card string, matchOn, include clientmodel.LabelNames) (*vectorMatching, error) {
var matchCardinalities = map[string]ast.VectorMatchCardinality{
"": ast.MatchOneToOne,
"GROUP_LEFT": ast.MatchManyToOne,
"GROUP_RIGHT": ast.MatchOneToMany,
}
matchCard, ok := matchCardinalities[card]
if !ok {
return nil, fmt.Errorf("invalid vector match cardinality %q", card)
}
if matchCard != ast.MatchOneToOne && len(include) == 0 {
return nil, fmt.Errorf("grouped vector matching must provide labels")
}
// There must be no overlap between both labelname lists.
for _, matchLabel := range matchOn {
for _, incLabel := range include {
if matchLabel == incLabel {
return nil, fmt.Errorf("use of label %s in ON and %s clauses not allowed", incLabel, card)
}
}
}
return &vectorMatching{matchCard, matchOn, include}, nil
}
// NewArithExpr is a convenience function to create a new AST arithmetic expression.
func NewArithExpr(opTypeStr string, lhs ast.Node, rhs ast.Node, vecMatching *vectorMatching) (ast.Node, error) {
var opTypes = map[string]ast.BinOpType{
"+": ast.Add,
"-": ast.Sub,
"*": ast.Mul,
"/": ast.Div,
"%": ast.Mod,
">": ast.GT,
"<": ast.LT,
"==": ast.EQ,
"!=": ast.NE,
">=": ast.GE,
"<=": ast.LE,
"AND": ast.And,
"OR": ast.Or,
}
opType, ok := opTypes[opTypeStr]
if !ok {
return nil, fmt.Errorf("invalid binary operator %q", opTypeStr)
}
var vm vectorMatching
if vecMatching != nil {
vm = *vecMatching
// And/or always do many-to-many matching.
if opType == ast.And || opType == ast.Or {
vm.matchCardinality = ast.MatchManyToMany
}
}
expr, err := ast.NewArithExpr(opType, lhs, rhs, vm.matchCardinality, vm.matchOn, vm.includeLabels)
if err != nil {
return nil, fmt.Errorf(err.Error())
}
return expr, nil
}
Implement offset operator. This allows changing the time offset for individual instant and range vectors in a query. For example, this returns the value of `foo` 5 minutes in the past relative to the current query evaluation time: foo offset 5m Note that the `offset` modifier always needs to follow the selector immediately. I.e. the following would be correct: sum(foo offset 5m) // GOOD. While the following would be *incorrect*: sum(foo) offset 5m // INVALID. The same works for range vectors. This returns the 5-minutes-rate that `foo` had a week ago: rate(foo[5m] offset 1w) This change touches the following components: * Lexer/parser: additions to correctly parse the new `offset`/`OFFSET` keyword. * AST: vector and matrix nodes now have an additional `offset` field. This is used during their evaluation to adjust query and result times appropriately. * Query analyzer: now works on separate sets of ranges and instants per offset. Isolating different offsets from each other completely in this way keeps the preloading code relatively simple. No storage engine changes were needed by this change. The rules tests have been changed to not probe the internal implementation details of the query analyzer anymore (how many instants and ranges have been preloaded). This would also become too cumbersome to test with the new model, and measuring the result of the query should be sufficient. This fixes https://github.com/prometheus/prometheus/issues/529 This fixed https://github.com/prometheus/promdash/issues/201
10 years ago
// NewVectorSelector is a convenience function to create a new AST vector selector.
func NewVectorSelector(m metric.LabelMatchers, offsetStr string) (ast.VectorNode, error) {
offset, err := utility.StringToDuration(offsetStr)
if err != nil {
return nil, err
}
Implement offset operator. This allows changing the time offset for individual instant and range vectors in a query. For example, this returns the value of `foo` 5 minutes in the past relative to the current query evaluation time: foo offset 5m Note that the `offset` modifier always needs to follow the selector immediately. I.e. the following would be correct: sum(foo offset 5m) // GOOD. While the following would be *incorrect*: sum(foo) offset 5m // INVALID. The same works for range vectors. This returns the 5-minutes-rate that `foo` had a week ago: rate(foo[5m] offset 1w) This change touches the following components: * Lexer/parser: additions to correctly parse the new `offset`/`OFFSET` keyword. * AST: vector and matrix nodes now have an additional `offset` field. This is used during their evaluation to adjust query and result times appropriately. * Query analyzer: now works on separate sets of ranges and instants per offset. Isolating different offsets from each other completely in this way keeps the preloading code relatively simple. No storage engine changes were needed by this change. The rules tests have been changed to not probe the internal implementation details of the query analyzer anymore (how many instants and ranges have been preloaded). This would also become too cumbersome to test with the new model, and measuring the result of the query should be sufficient. This fixes https://github.com/prometheus/prometheus/issues/529 This fixed https://github.com/prometheus/promdash/issues/201
10 years ago
return ast.NewVectorSelector(m, offset), nil
}
// NewMatrixSelector is a convenience function to create a new AST matrix selector.
func NewMatrixSelector(vector ast.Node, intervalStr string, offsetStr string) (ast.MatrixNode, error) {
interval, err := utility.StringToDuration(intervalStr)
if err != nil {
return nil, err
}
Implement offset operator. This allows changing the time offset for individual instant and range vectors in a query. For example, this returns the value of `foo` 5 minutes in the past relative to the current query evaluation time: foo offset 5m Note that the `offset` modifier always needs to follow the selector immediately. I.e. the following would be correct: sum(foo offset 5m) // GOOD. While the following would be *incorrect*: sum(foo) offset 5m // INVALID. The same works for range vectors. This returns the 5-minutes-rate that `foo` had a week ago: rate(foo[5m] offset 1w) This change touches the following components: * Lexer/parser: additions to correctly parse the new `offset`/`OFFSET` keyword. * AST: vector and matrix nodes now have an additional `offset` field. This is used during their evaluation to adjust query and result times appropriately. * Query analyzer: now works on separate sets of ranges and instants per offset. Isolating different offsets from each other completely in this way keeps the preloading code relatively simple. No storage engine changes were needed by this change. The rules tests have been changed to not probe the internal implementation details of the query analyzer anymore (how many instants and ranges have been preloaded). This would also become too cumbersome to test with the new model, and measuring the result of the query should be sufficient. This fixes https://github.com/prometheus/prometheus/issues/529 This fixed https://github.com/prometheus/promdash/issues/201
10 years ago
offset, err := utility.StringToDuration(offsetStr)
if err != nil {
return nil, err
}
vectorSelector, ok := vector.(*ast.VectorSelector)
if !ok {
return nil, fmt.Errorf("intervals are currently only supported for vector selectors")
}
return ast.NewMatrixSelector(vectorSelector, interval, offset), nil
}
func newLabelMatcher(matchTypeStr string, name clientmodel.LabelName, value clientmodel.LabelValue) (*metric.LabelMatcher, error) {
matchTypes := map[string]metric.MatchType{
"=": metric.Equal,
"!=": metric.NotEqual,
"=~": metric.RegexMatch,
"!~": metric.RegexNoMatch,
}
matchType, ok := matchTypes[matchTypeStr]
if !ok {
return nil, fmt.Errorf("invalid label matching operator %q", matchTypeStr)
}
return metric.NewLabelMatcher(matchType, name, value)
}
// TableLinkForExpression creates an escaped relative link to the table view of
// the provided expression.
func TableLinkForExpression(expr string) string {
// url.QueryEscape percent-escapes everything except spaces, for which it
// uses "+". However, in the non-query part of a URI, only percent-escaped
// spaces are legal, so we need to manually replace "+" with "%20" after
// query-escaping the string.
//
// See also:
// http://stackoverflow.com/questions/1634271/url-encoding-the-space-character-or-20.
urlData := url.QueryEscape(fmt.Sprintf(`[{"expr":%q,"tab":1}]`, expr))
return fmt.Sprintf("/graph#%s", strings.Replace(urlData, "+", "%20", -1))
}
// GraphLinkForExpression creates an escaped relative link to the graph view of
// the provided expression.
func GraphLinkForExpression(expr string) string {
urlData := url.QueryEscape(fmt.Sprintf(`[{"expr":%q,"tab":0}]`, expr))
return fmt.Sprintf("/graph#%s", strings.Replace(urlData, "+", "%20", -1))
}