-
Notifications
You must be signed in to change notification settings - Fork 309
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
lwc-events: improve data point conversion (#1631)
Updates to support common statistics for timer and dist summary. Aggregates the values locally to reduce number of messages.
- Loading branch information
1 parent
3501fc8
commit fc80466
Showing
6 changed files
with
601 additions
and
62 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
243 changes: 243 additions & 0 deletions
243
atlas-lwc-events/src/main/scala/com/netflix/atlas/lwc/events/DatapointConverter.scala
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,243 @@ | ||
| /* | ||
| * Copyright 2014-2024 Netflix, Inc. | ||
| * | ||
| * 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 com.netflix.atlas.lwc.events | ||
|
|
||
| import com.netflix.atlas.core.model.DataExpr | ||
| import com.netflix.atlas.core.model.Query | ||
| import com.netflix.spectator.api.Clock | ||
| import com.netflix.spectator.impl.AtomicDouble | ||
| import com.netflix.spectator.impl.StepDouble | ||
|
|
||
| import java.util.concurrent.ConcurrentHashMap | ||
|
|
||
| /** | ||
| * Helper to convert a sequence of events into a data point. | ||
| */ | ||
| private[events] trait DatapointConverter { | ||
|
|
||
| def update(event: LwcEvent): Unit | ||
|
|
||
| def flush(timestamp: Long): Unit | ||
| } | ||
|
|
||
| private[events] object DatapointConverter { | ||
|
|
||
| def apply( | ||
| id: String, | ||
| expr: DataExpr, | ||
| clock: Clock, | ||
| step: Long, | ||
| consumer: (String, LwcEvent) => Unit | ||
| ): DatapointConverter = { | ||
| val tags = Query.tags(expr.query) | ||
| val mapper = createValueMapper(tags, expr.finalGrouping) | ||
| val params = Params(id, Query.tags(expr.query), clock, step, mapper, consumer) | ||
| toConverter(expr, params) | ||
| } | ||
|
|
||
| private def toConverter(expr: DataExpr, params: Params): DatapointConverter = { | ||
| expr match { | ||
| case _: DataExpr.Sum => Sum(params) | ||
| case _: DataExpr.Count => Count(params) | ||
| case _: DataExpr.Max => Max(params) | ||
| case _: DataExpr.Min => Min(params) | ||
| case by: DataExpr.GroupBy => GroupBy(by, params) | ||
| case _ => Sum(params) | ||
| } | ||
| } | ||
|
|
||
| /** | ||
| * Extract value and map as needed based on the type. Uses statistic and grouping to | ||
| * coerce events so they structurally work like spectator composite types. | ||
| */ | ||
| def createValueMapper(tags: Map[String, String], grouping: List[String]): LwcEvent => Double = { | ||
| tags.get("value") match { | ||
| case Some(k) => | ||
| tags.get("statistic") match { | ||
| case Some("count") => _ => 1.0 | ||
| case Some("totalOfSquares") => event => squared(event.extractValue(k)) | ||
| case _ => event => toDouble(event.extractValue(k)) | ||
| } | ||
| case None => | ||
| _ => 1.0 | ||
| } | ||
| } | ||
|
|
||
| private def squared(value: Any): Double = { | ||
| val v = toDouble(value) | ||
| v * v | ||
| } | ||
|
|
||
| def toDouble(value: Any): Double = { | ||
| value match { | ||
| case v: Boolean => if (v) 1.0 else 0.0 | ||
| case v: Byte => v.toDouble | ||
| case v: Short => v.toDouble | ||
| case v: Int => v.toDouble | ||
| case v: Long => v.toDouble | ||
| case v: Float => v.toDouble | ||
| case v: Double => v | ||
| case v: Number => v.doubleValue() | ||
| case v: String => parseDouble(v) | ||
| case _ => 1.0 | ||
| } | ||
| } | ||
|
|
||
| private def parseDouble(v: String): Double = { | ||
| try { | ||
| java.lang.Double.parseDouble(v) | ||
| } catch { | ||
| case _: Exception => Double.NaN | ||
| } | ||
| } | ||
|
|
||
| case class Params( | ||
| id: String, | ||
| tags: Map[String, String], | ||
| clock: Clock, | ||
| step: Long, | ||
| valueMapper: LwcEvent => Double, | ||
| consumer: (String, LwcEvent) => Unit | ||
| ) { | ||
|
|
||
| val buffer = new StepDouble(0.0, clock, step) | ||
| } | ||
|
|
||
| /** Compute sum for a counter as a rate per second. */ | ||
| case class Sum(params: Params) extends DatapointConverter { | ||
|
|
||
| override def update(event: LwcEvent): Unit = { | ||
| val value = params.valueMapper(event) | ||
| if (value.isFinite && value >= 0.0) { | ||
| params.buffer.getCurrent.addAndGet(value) | ||
| } | ||
| } | ||
|
|
||
| override def flush(timestamp: Long): Unit = { | ||
| val value = params.buffer.pollAsRate(timestamp) | ||
| if (value.isFinite) { | ||
| val ts = timestamp / params.step * params.step | ||
| val event = DatapointEvent(params.id, params.tags, ts, value) | ||
| params.consumer(params.id, event) | ||
| } | ||
| } | ||
| } | ||
|
|
||
| /** Compute count of contributing events. */ | ||
| case class Count(params: Params) extends DatapointConverter { | ||
|
|
||
| override def update(event: LwcEvent): Unit = { | ||
| params.buffer.getCurrent.addAndGet(1.0) | ||
| } | ||
|
|
||
| override def flush(timestamp: Long): Unit = { | ||
| val value = params.buffer.poll(timestamp) | ||
| if (value.isFinite) { | ||
| val ts = timestamp / params.step * params.step | ||
| val event = DatapointEvent(params.id, params.tags, ts, value) | ||
| params.consumer(params.id, event) | ||
| } | ||
| } | ||
| } | ||
|
|
||
| /** Compute max value from contributing events. */ | ||
| case class Max(params: Params) extends DatapointConverter { | ||
|
|
||
| override def update(event: LwcEvent): Unit = { | ||
| val value = params.valueMapper(event) | ||
| if (value.isFinite && value >= 0.0) { | ||
| params.buffer.getCurrent.max(value) | ||
| } | ||
| } | ||
|
|
||
| override def flush(timestamp: Long): Unit = { | ||
| val value = params.buffer.poll(timestamp) | ||
| if (value.isFinite) { | ||
| val ts = timestamp / params.step * params.step | ||
| val event = DatapointEvent(params.id, params.tags, ts, value) | ||
| params.consumer(params.id, event) | ||
| } | ||
| } | ||
| } | ||
|
|
||
| /** Compute min value from contributing events. */ | ||
| case class Min(params: Params) extends DatapointConverter { | ||
|
|
||
| override def update(event: LwcEvent): Unit = { | ||
| val value = params.valueMapper(event) | ||
| if (value.isFinite && value >= 0.0) { | ||
| min(params.buffer.getCurrent, value) | ||
| } | ||
| } | ||
|
|
||
| private def min(current: AtomicDouble, value: Double): Unit = { | ||
| if (value.isFinite) { | ||
| var min = current.get() | ||
| while (isLessThan(value, min) && !current.compareAndSet(min, value)) { | ||
| min = current.get() | ||
| } | ||
| } | ||
| } | ||
|
|
||
| private def isLessThan(v1: Double, v2: Double): Boolean = { | ||
| v1 < v2 || v2.isNaN | ||
| } | ||
|
|
||
| override def flush(timestamp: Long): Unit = { | ||
| val value = params.buffer.poll(timestamp) | ||
| if (value.isFinite) { | ||
| val ts = timestamp / params.step * params.step | ||
| val event = DatapointEvent(params.id, params.tags, ts, value) | ||
| params.consumer(params.id, event) | ||
| } | ||
| } | ||
| } | ||
|
|
||
| /** Compute set of data points, one for each distinct group. */ | ||
| case class GroupBy(by: DataExpr.GroupBy, params: Params) extends DatapointConverter { | ||
|
|
||
| private val groups = new ConcurrentHashMap[Map[String, String], DatapointConverter]() | ||
|
|
||
| override def update(event: LwcEvent): Unit = { | ||
| // Ignore events that are missing dimensions used in the grouping | ||
| val values = by.keys.map(event.tagValue).filterNot(_ == null) | ||
| if (by.keys.size == values.size) { | ||
| val value = params.valueMapper(event) | ||
| if (value.isFinite) { | ||
| val tags = by.keys.zip(values).toMap | ||
| val converter = groups.computeIfAbsent(tags, groupConverter) | ||
| converter.update(event) | ||
| } | ||
| } | ||
| } | ||
|
|
||
| private def groupConverter(tags: Map[String, String]): DatapointConverter = { | ||
| val ps = params.copy(consumer = (id, event) => groupConsumer(tags, id, event)) | ||
| toConverter(by.af, ps) | ||
| } | ||
|
|
||
| private def groupConsumer(tags: Map[String, String], id: String, event: LwcEvent): Unit = { | ||
| event match { | ||
| case dp: DatapointEvent => params.consumer(id, dp.copy(tags = dp.tags ++ tags)) | ||
| case ev => params.consumer(id, ev) | ||
| } | ||
| } | ||
|
|
||
| override def flush(timestamp: Long): Unit = { | ||
| groups.values().forEach(_.flush(timestamp)) | ||
| } | ||
| } | ||
| } |
Oops, something went wrong.