Github user aledsage commented on a diff in the pull request:
https://github.com/apache/brooklyn-server/pull/908#discussion_r154934710
--- Diff:
core/src/main/java/org/apache/brooklyn/enricher/stock/aggregator/AggregationJob.java
---
@@ -0,0 +1,158 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you 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 org.apache.brooklyn.enricher.stock.aggregator;
+
+import java.util.ArrayList;
+import java.util.HashMap;
+import java.util.Iterator;
+import java.util.List;
+import java.util.Map;
+
+import org.apache.commons.lang3.StringUtils;
+
+import org.apache.brooklyn.api.entity.Entity;
+import org.apache.brooklyn.core.sensor.BasicAttributeSensorAndConfigKey;
+
+public final class AggregationJob implements Runnable {
+
+ public static BasicAttributeSensorAndConfigKey<Map<String, String>>
DASHBOARD_COST_PER_MONTH = new BasicAttributeSensorAndConfigKey(Map.class,
"dashboard.costPerMonth", "A map of VM ID to monthly cost. Please note if the
same VM ID is reported more than once, only one of the sensors will be
propagated");
+ public static BasicAttributeSensorAndConfigKey<List<Map<String,
String>>> DASHBOARD_HA_PRIMARIES = new
BasicAttributeSensorAndConfigKey(List.class, "dashboard.ha.primaries", "A list
of high availability primaries "); //TODO: find out what this is and add a
better description
+ public static BasicAttributeSensorAndConfigKey<List<Map<String,
String>>> DASHBOARD_LICENSES = new BasicAttributeSensorAndConfigKey(List.class,
"dashboard.licenses", "The licences in use for the running entities. This is a
list of maps. The map should contain two keys: product and key");
+ public static BasicAttributeSensorAndConfigKey<Map<String,
List<Map<String, Object>>>> DASHBOARD_LOCATIONS = new
BasicAttributeSensorAndConfigKey(Map.class, "dashboard.locations", "Locations
in which the VMs are running. A map where the key is category of location (e.g.
server) and the value is a list of maps containing name/icon/count");
+ public static BasicAttributeSensorAndConfigKey<List<Map<String,
String>>> DASHBOARD_POLICY_HIGHLIGHTS = new
BasicAttributeSensorAndConfigKey(List.class, "dashboard.policyHighlights",
"Highlights from policies. List of Masps, where each map should contain text
and category");
+
+ private final Entity entity;
+
+ private static final String NAME_STRING = "name";
+ private static final String COUNT_STRING = "count";
+
+ public AggregationJob(Entity entity) {
+ this.entity = entity;
+ }
+
+ private void updateFromConfig(Entity entity,
BasicAttributeSensorAndConfigKey<List<Map<String, String>>> sensorAndConfigKey,
List<Map<String, String>> list) {
+ List<Map<String, String>> listFromEntity =
entity.sensors().get(sensorAndConfigKey);
+ if (listFromEntity == null || listFromEntity.isEmpty()) {
+ listFromEntity = entity.config().get(sensorAndConfigKey);
+ }
+
+ if (listFromEntity != null) {
+ list.addAll(listFromEntity);
+ }
+ }
+
+ @Override
+ public void run() {
+ HashMap<String, String> costPerMonth = new HashMap<>();
+ List<Map<String, String>> haPrimaries = new ArrayList<>();
+ List<Map<String, String>> dashboardLicences = new ArrayList<>();
+ Map<String, List<Map<String, Object>>> dashboardLocations = new
HashMap<>();
+ List<Map<String, String>> dashboardPolicyHighlights = new
ArrayList<>();
+
+ entity.getChildren().forEach(childEntity ->
scanEntity(childEntity, costPerMonth, haPrimaries, dashboardLicences,
dashboardLocations, dashboardPolicyHighlights));
+ entity.sensors().set(DASHBOARD_COST_PER_MONTH, costPerMonth);
+ entity.sensors().set(DASHBOARD_HA_PRIMARIES, haPrimaries);
+ entity.sensors().set(DASHBOARD_LICENSES, dashboardLicences);
+ entity.sensors().set(DASHBOARD_LOCATIONS, dashboardLocations);
+ entity.sensors().set(DASHBOARD_POLICY_HIGHLIGHTS,
dashboardPolicyHighlights);
+ }
+
+ protected void scanEntity(Entity entity,
+ Map<String, String> costPerMonth,
+ List<Map<String, String>> haPrimaries,
+ List<Map<String, String>> licences,
+ Map<String, List<Map<String, Object>>>
locations,
+ List<Map<String, String>> policyHighlights) {
+
+ updateFromConfig(entity, DASHBOARD_HA_PRIMARIES, haPrimaries);
+ updateFromConfig(entity, DASHBOARD_LICENSES, licences);
+ updateFromConfig(entity, DASHBOARD_POLICY_HIGHLIGHTS,
policyHighlights);
+
+ //Cost per month
+ Map<String, String> costPerMonthFromEntity =
entity.sensors().get(DASHBOARD_COST_PER_MONTH);
+ if (costPerMonthFromEntity == null ||
costPerMonthFromEntity.isEmpty()) {
+ costPerMonthFromEntity =
entity.config().get(DASHBOARD_COST_PER_MONTH);
+ }
+
+ if (costPerMonthFromEntity != null) {
+ costPerMonth.putAll(costPerMonthFromEntity);
+ }
+
+ //Locations merge
+ //We are merging a Map, of Lists of Maps
+ //The outer map is location type, e.g. servers to a list of data
for that type
+ //The inner map should contain the data e.g. name/icon/count.
+
+ //Further complicating this is that the Map you get back from
sensors/config doesn't conform to the generic expectations.
+ //I.E. even if you type this as Map<String, List<Map<String,
String>>> you will still get back an integer from
+ //the inner map- hence all the weird casting bellow.
+ Map<String, List<Map<String, Object>>> outerLocationMapFromEntity
= entity.sensors().get(DASHBOARD_LOCATIONS);
+ if (outerLocationMapFromEntity == null ||
outerLocationMapFromEntity.isEmpty()) {
+ outerLocationMapFromEntity =
entity.config().get(DASHBOARD_LOCATIONS);
+ }
+
+ if (outerLocationMapFromEntity != null) {
+ //loop through outer maps
+
outerLocationMapFromEntity.forEach((outerLocationMapFromEntityKey,
outerLocationMapFromEntityValue) -> {
+ boolean found = false;
+ for (Map.Entry<String, List<Map<String, Object>>>
outerLocationMapFromMethodParam : locations.entrySet()) {
+ if
(StringUtils.equals(outerLocationMapFromMethodParam.getKey(),
outerLocationMapFromEntityKey)) {
+ found = true;
+
+ //loop through list
+ Iterator<Map<String, Object>>
listIteratorFromEntity = outerLocationMapFromEntityValue.iterator();
+ while (listIteratorFromEntity.hasNext()) {
+ Map<String, Object> innerMapFromEntity =
listIteratorFromEntity.next();
+ boolean foundInner = false;
+
+ //loop through inner map and merge
+ for (Map<String, Object>
innerMapFromMethodParam : outerLocationMapFromMethodParam.getValue()) {
+ if (StringUtils.equals((String)
innerMapFromEntity.get(NAME_STRING), (String)
innerMapFromMethodParam.get(NAME_STRING))) {
+
+
innerMapFromMethodParam.put(COUNT_STRING, (int)
innerMapFromEntity.get(COUNT_STRING) + (int)
innerMapFromMethodParam.get(COUNT_STRING));
+ foundInner = true;
+ break;
+ }
+ }
+
+ //If the entity has a "name" not found in the
method param then add it to the method param
+ if (!foundInner) {
+
outerLocationMapFromMethodParam.getValue().add(new
HashMap<>(innerMapFromEntity));
+ }
+ }
+ }
+
+ }
+
+ //If the entity has an entry in the outer map that isn't
in the method param, then add it
+ if (!found) {
+ ArrayList clonedList = new ArrayList();
+ outerLocationMapFromEntityValue.forEach(mapToClone -> {
+ clonedList.add(new HashMap<>(mapToClone));
+
+ });
+ locations.put(outerLocationMapFromEntityKey,
clonedList);
+ }
+
+ });
+ }
+
+ entity.getChildren().forEach(childEntity ->
scanEntity(childEntity, costPerMonth, haPrimaries, licences, locations,
policyHighlights));
--- End diff --
The recursion here feels strange - if we have a cluster with a
`costPerMonth` sensor, would you expect that to already be an aggregate of its
children? Or would you expect to add that and also look at all its children to
add each of those individually as well?
I think you could argue it either way - it will sometimes feel surprising
no matter which way round you choose (e.g. if I have a software process with
child software processes - e.g. where the child installs a new relic agent -
then I'd expect to take license info from the parent software process and the
child software process).
---
