-
Notifications
You must be signed in to change notification settings - Fork 24.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Optimize usage calculation in ILM policies retrieval API #106953
base: main
Are you sure you want to change the base?
Changes from 8 commits
5bf393c
d160dc2
092cb0c
16f3d83
d30376e
24acf2c
293ff9d
02fb7a2
e193a30
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,6 @@ | ||
pr: 106953 | ||
summary: Optimize usage calculation in ILM policies retrieval API | ||
area: ILM+SLM | ||
type: enhancement | ||
issues: | ||
- 105773 |
Original file line number | Diff line number | Diff line change |
---|---|---|
|
@@ -61,6 +61,7 @@ | |
import java.time.Instant; | ||
import java.util.ArrayList; | ||
import java.util.Arrays; | ||
import java.util.Collection; | ||
import java.util.Collections; | ||
import java.util.Comparator; | ||
import java.util.HashMap; | ||
|
@@ -1050,29 +1051,25 @@ static Set<String> dataStreamsExclusivelyUsingTemplates(final ClusterState state | |
.reduce(Sets::union) | ||
.orElse(Set.of()); | ||
|
||
// Determine all the composable templates that are not one of the provided templates. | ||
var otherTemplates = state.metadata() | ||
.templatesV2() | ||
.entrySet() | ||
.stream() | ||
.filter( | ||
entry -> templateNames.contains(entry.getKey()) == false | ||
&& isGlobalAndHasIndexHiddenSetting(metadata, entry.getValue(), entry.getKey()) == false | ||
) | ||
// Sort here so we can `exitOnFirstMatch` in `findV2Template`. | ||
.sorted(Comparator.comparing(entry -> entry.getValue().priorityOrZero(), Comparator.reverseOrder())) | ||
.toList(); | ||
|
||
return metadata.dataStreams() | ||
.values() | ||
.stream() | ||
// Limit to checking data streams that match any of the templates' index patterns | ||
.filter(ds -> namePatterns.stream().anyMatch(pattern -> Regex.simpleMatch(pattern, ds.getName()))) | ||
.filter(ds -> { | ||
// Retrieve the templates that match the data stream name ordered by priority | ||
List<Tuple<String, ComposableIndexTemplate>> candidates = findV2CandidateTemplates(metadata, ds.getName(), ds.isHidden()); | ||
if (candidates.isEmpty()) { | ||
throw new IllegalStateException("Data stream " + ds.getName() + " did not match any composable index templates."); | ||
} | ||
|
||
// Limit data streams that can ONLY use any of the specified templates, we do this by filtering | ||
// the matching templates that are others than the ones requested and could be a valid template to use. | ||
return candidates.stream() | ||
.filter( | ||
template -> templateNames.contains(template.v1()) == false | ||
&& isGlobalAndHasIndexHiddenSetting(metadata, template.v2(), template.v1()) == false | ||
) | ||
.map(Tuple::v1) | ||
.toList() | ||
.isEmpty(); | ||
}) | ||
.filter(ds -> findV2Template(state.metadata(), otherTemplates, ds.getName(), ds.isHidden(), true) == null) | ||
.map(DataStream::getName) | ||
.collect(Collectors.toSet()); | ||
} | ||
|
@@ -1268,7 +1265,27 @@ public static List<IndexTemplateMetadata> findV1Templates(Metadata metadata, Str | |
*/ | ||
@Nullable | ||
public static String findV2Template(Metadata metadata, String indexName, boolean isHidden) { | ||
final List<Tuple<String, ComposableIndexTemplate>> candidates = findV2CandidateTemplates(metadata, indexName, isHidden); | ||
return findV2Template(metadata, metadata.templatesV2().entrySet(), indexName, isHidden, false); | ||
} | ||
Comment on lines
1267
to
+1269
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. What if we create one more method called: I think this will clarify a bit more which method to use and it will make explicit that the templates need to be ordered based on priority. |
||
|
||
/** | ||
* Return the name (id) of the highest matching index template, out of the provided templates, for the given index name. In | ||
* the event that no templates are matched, {@code null} is returned. | ||
*/ | ||
@Nullable | ||
public static String findV2Template( | ||
Metadata metadata, | ||
Collection<Map.Entry<String, ComposableIndexTemplate>> templates, | ||
String indexName, | ||
boolean isHidden, | ||
boolean exitOnFirstMatch | ||
) { | ||
final List<Tuple<String, ComposableIndexTemplate>> candidates = findV2CandidateTemplates( | ||
templates, | ||
indexName, | ||
isHidden, | ||
exitOnFirstMatch | ||
); | ||
Comment on lines
+1276
to
+1288
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. If you like the idea of more explicit method names like I suggested in the previous comment. We could change the visibility of this method to private or package private depending on the testing requirements. This way this method won't be misused by setting What do you think? |
||
if (candidates.isEmpty()) { | ||
return null; | ||
} | ||
|
@@ -1296,25 +1313,30 @@ public static String findV2Template(Metadata metadata, String indexName, boolean | |
/** | ||
* Return an ordered list of the name (id) and composable index templates that would apply to an index. The first | ||
* one is the winner template that is applied to this index. In the event that no templates are matched, | ||
* an empty list is returned. | ||
* an empty list is returned. If <code>exitOnFirstMatch</code> is true, we return immediately after finding a match. | ||
*/ | ||
Comment on lines
1314
to
1317
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. I believe we need to elaborate a bit more on how to use this method. What are the trade-offs of |
||
static List<Tuple<String, ComposableIndexTemplate>> findV2CandidateTemplates(Metadata metadata, String indexName, boolean isHidden) { | ||
static List<Tuple<String, ComposableIndexTemplate>> findV2CandidateTemplates( | ||
Collection<Map.Entry<String, ComposableIndexTemplate>> templates, | ||
String indexName, | ||
boolean isHidden, | ||
boolean exitOnFirstMatch | ||
) { | ||
final String resolvedIndexName = IndexNameExpressionResolver.DateMathExpressionResolver.resolveExpression(indexName); | ||
final Predicate<String> patternMatchPredicate = pattern -> Regex.simpleMatch(pattern, resolvedIndexName); | ||
final List<Tuple<String, ComposableIndexTemplate>> candidates = new ArrayList<>(); | ||
for (Map.Entry<String, ComposableIndexTemplate> entry : metadata.templatesV2().entrySet()) { | ||
for (Map.Entry<String, ComposableIndexTemplate> entry : templates) { | ||
final String name = entry.getKey(); | ||
final ComposableIndexTemplate template = entry.getValue(); | ||
if (isHidden == false) { | ||
final boolean matched = template.indexPatterns().stream().anyMatch(patternMatchPredicate); | ||
if (matched) { | ||
candidates.add(Tuple.tuple(name, template)); | ||
if (isHidden) { | ||
final boolean hasMatchAllTemplate = template.indexPatterns().stream().anyMatch(Regex::isMatchAllPattern); | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Nit: |
||
if (hasMatchAllTemplate) { | ||
continue; | ||
} | ||
} else { | ||
final boolean isNotMatchAllTemplate = template.indexPatterns().stream().noneMatch(Regex::isMatchAllPattern); | ||
if (isNotMatchAllTemplate) { | ||
if (template.indexPatterns().stream().anyMatch(patternMatchPredicate)) { | ||
candidates.add(Tuple.tuple(name, template)); | ||
} | ||
for (String indexPattern : template.indexPatterns()) { | ||
if (Regex.simpleMatch(indexPattern, resolvedIndexName)) { | ||
candidates.add(Tuple.tuple(name, template)); | ||
if (exitOnFirstMatch) { | ||
return candidates; | ||
} | ||
nielsbauman marked this conversation as resolved.
Show resolved
Hide resolved
|
||
} | ||
} | ||
|
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,105 @@ | ||
/* | ||
* Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one | ||
* or more contributor license agreements. Licensed under the Elastic License | ||
* 2.0; you may not use this file except in compliance with the Elastic License | ||
* 2.0. | ||
*/ | ||
|
||
package org.elasticsearch.xpack.core.ilm; | ||
|
||
import org.apache.lucene.util.CollectionUtil; | ||
import org.elasticsearch.action.support.IndicesOptions; | ||
import org.elasticsearch.cluster.ClusterState; | ||
import org.elasticsearch.cluster.metadata.ComposableIndexTemplate; | ||
import org.elasticsearch.cluster.metadata.IndexMetadata; | ||
import org.elasticsearch.cluster.metadata.IndexNameExpressionResolver; | ||
import org.elasticsearch.cluster.metadata.ItemUsage; | ||
import org.elasticsearch.cluster.metadata.MetadataIndexTemplateService; | ||
import org.elasticsearch.common.regex.Regex; | ||
import org.elasticsearch.common.settings.Settings; | ||
|
||
import java.util.ArrayList; | ||
import java.util.Comparator; | ||
import java.util.HashMap; | ||
import java.util.List; | ||
import java.util.Map; | ||
|
||
/** | ||
* A class that can be used to calculate the usages of ILM policies. This class computes some information on initialization, which will | ||
* use a bit more memory but speeds up the usage calculation significantly. | ||
*/ | ||
public class LifecyclePolicyUsageCalculator { | ||
|
||
private final ClusterState state; | ||
/** Whether {@link #calculateUsage} will be called multiple times or not. */ | ||
private final boolean willIterate; | ||
/** A map from policy name to list of data streams that use that policy. */ | ||
private final Map<String, List<String>> policyToDataStream; | ||
/** A map from composable template name to the policy name it uses (or null) */ | ||
private final Map<String, String> templateToPolicy; | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. The |
||
|
||
public LifecyclePolicyUsageCalculator( | ||
final IndexNameExpressionResolver indexNameExpressionResolver, | ||
final ClusterState state, | ||
List<String> names | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Because we have a lot of elements involved here, I think it would be easier to specify what are these What do you think? |
||
) { | ||
this.state = state; | ||
this.willIterate = names.size() > 1 || Regex.isSimpleMatchPattern(names.get(0)); | ||
|
||
var allDataStreams = indexNameExpressionResolver.dataStreamNames(state, IndicesOptions.LENIENT_EXPAND_OPEN_CLOSED_HIDDEN); | ||
// Sort all templates by descending priority. That way, findV2Template can exit on the first found template. | ||
var indexTemplates = new ArrayList<>(state.metadata().templatesV2().entrySet()); | ||
CollectionUtil.timSort(indexTemplates, Comparator.comparing(entry -> entry.getValue().priorityOrZero(), Comparator.reverseOrder())); | ||
|
||
// Build the maps that will be used for the usage calculation later on. | ||
IndexLifecycleMetadata metadata = state.metadata().custom(IndexLifecycleMetadata.TYPE); | ||
policyToDataStream = new HashMap<>(Regex.isSimpleMatchPattern(names.get(0)) ? metadata.getPolicyMetadatas().size() : names.size()); | ||
templateToPolicy = new HashMap<>(indexTemplates.size()); | ||
for (String dataStream : allDataStreams) { | ||
String indexTemplate = MetadataIndexTemplateService.findV2Template(state.metadata(), indexTemplates, dataStream, false, true); | ||
if (indexTemplate == null) { | ||
continue; | ||
} | ||
Settings settings = MetadataIndexTemplateService.resolveSettings(state.metadata(), indexTemplate); | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Before doing this, shouldn't we check if the template is already in |
||
var policyName = LifecycleSettings.LIFECYCLE_NAME_SETTING.get(settings); | ||
if (names.stream().noneMatch(name -> Regex.simpleMatch(name, policyName))) { | ||
// If a template's policy doesn't match any of the supplied names, we can skip it later on. | ||
templateToPolicy.put(indexTemplate, null); | ||
continue; | ||
} | ||
templateToPolicy.put(indexTemplate, policyName); | ||
policyToDataStream.computeIfAbsent(policyName, k -> new ArrayList<>()).add(dataStream); | ||
} | ||
} | ||
|
||
/** | ||
* Calculate the indices, data streams, and composable templates that use the given policy. | ||
*/ | ||
public ItemUsage calculateUsage(String policyName) { | ||
List<String> indices = new ArrayList<>(); | ||
for (IndexMetadata indexMetadata : state.metadata().indices().values()) { | ||
if (policyName.equals(indexMetadata.getLifecyclePolicyName())) { | ||
indices.add(indexMetadata.getIndex().getName()); | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. I could pre-compute a map of policy to indices as well, but I'm not sure the memory vs. speed trade-off is worth it there. There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. On second thought, since we have to build this list for every policy anyway, pre-computing these lists and putting them in a map shouldn't be too much additional memory overhead. I'll wait till someone has done a first review before making more changes (in case my whole approach is off). There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. And the same goes for the composable templates of course. There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Let's think about this. You only need to keep a cache of what is necessary. Right? For example: Data streams Indices Then retrieving the data is just picking them up from the cache. Thoughts? |
||
} | ||
} | ||
|
||
List<String> composableTemplates = new ArrayList<>(); | ||
for (Map.Entry<String, ComposableIndexTemplate> entry : state.metadata().templatesV2().entrySet()) { | ||
var foundPolicy = templateToPolicy.get(entry.getKey()); | ||
// Extra `containsKey` check to account for templates not using any policy. | ||
if (foundPolicy == null && templateToPolicy.containsKey(entry.getKey()) == false) { | ||
Settings settings = MetadataIndexTemplateService.resolveSettings(entry.getValue(), state.metadata().componentTemplates()); | ||
foundPolicy = LifecycleSettings.LIFECYCLE_NAME_SETTING.get(settings); | ||
// If this method will only be called once, we don't need to keep building the map. | ||
if (willIterate) { | ||
templateToPolicy.put(entry.getKey(), foundPolicy); | ||
} | ||
} | ||
if (policyName.equals(foundPolicy)) { | ||
composableTemplates.add(entry.getKey()); | ||
} | ||
} | ||
|
||
return new ItemUsage(indices, policyToDataStream.getOrDefault(policyName, List.of()), composableTemplates); | ||
} | ||
} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Nit: I know this is not your code but since it's in my visual I wanted to add a small note.
I think this code is a bit more complex than it has to be.
Effectively we want to add all patterns into a set, right? I think the following code is a bit more readable.
What do you think?