Skip to content

Simplify EsqlExecution info serialization #131823

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.

Sign up for GitHub

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

Merged
merged 4 commits into from
Jul 28, 2025
Merged

Simplify EsqlExecution info serialization #131823

Changes from 1 commit
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
9b432fe
Simplify EsqlExecution info serialization
idegtiarenko Jul 24, 2025
626f7f6
avoid intermediate map
idegtiarenko Jul 25, 2025
69f1900
Merge branch 'main' into simplify_serialization
idegtiarenko Jul 25, 2025
3498968
Merge branch 'main' into simplify_serialization
idegtiarenko Jul 28, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
31 changes: 7 additions & 24 deletions x-pack/plugin/esql/src/main/java/org/elasticsearch/xpack/esql/action/EsqlExecutionInfo.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,6 @@
import org.elasticsearch.common.io.stream.StreamInput;
import org.elasticsearch.common.io.stream.StreamOutput;
import org.elasticsearch.common.io.stream.Writeable;
import org.elasticsearch.common.util.concurrent.ConcurrentCollections;
import org.elasticsearch.common.xcontent.ChunkedToXContentHelper;
import org.elasticsearch.common.xcontent.ChunkedToXContentObject;
import org.elasticsearch.core.Predicates;
Expand All @@ -37,6 +36,7 @@
import java.util.Map;
import java.util.Objects;
import java.util.Set;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.ConcurrentMap;
import java.util.function.BiFunction;
import java.util.function.Predicate;
Expand Down Expand Up @@ -66,7 +66,7 @@ public class EsqlExecutionInfo implements ChunkedToXContentObject, Writeable {
// The Map itself is immutable after construction - all Clusters will be accounted for at the start of the search.
// Updates to the Cluster occur with the updateCluster method that given the key to map transforms an
// old Cluster Object to a new Cluster Object with the remapping function.
public final Map<String, Cluster> clusterInfo;
public final ConcurrentMap<String, Cluster> clusterInfo;
// whether the user has asked for CCS metadata to be in the JSON response (the overall took will always be present)
private final boolean includeCCSMetadata;

Expand All @@ -90,7 +90,7 @@ public EsqlExecutionInfo(boolean includeCCSMetadata) {
* @param includeCCSMetadata (user defined setting) whether to include the CCS metadata in the HTTP response
*/
public EsqlExecutionInfo(Predicate<String> skipUnavailablePredicate, boolean includeCCSMetadata) {
this.clusterInfo = ConcurrentCollections.newConcurrentMap();
this.clusterInfo = new ConcurrentHashMap<>();
Copy link
Contributor

@smalyshev smalyshev Jul 24, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What is the difference here?

Copy link
Contributor Author

@idegtiarenko idegtiarenko Jul 25, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

elasticsearch/server/src/main/java/org/elasticsearch/common/util/concurrent/ConcurrentCollections.java

Lines 44 to 46 in a59c182

public static <K, V> ConcurrentMap<K, V> newConcurrentMap() {
return new ConcurrentHashMap<>();
}

newConcurrentMap simply delegates to new ConcurrentHashMap. Not sure why we have it. I am replacing it with what it actually is.

this.skipUnavailablePredicate = skipUnavailablePredicate;
this.includeCCSMetadata = includeCCSMetadata;
this.relativeStart = TimeSpan.start();
Expand All @@ -108,26 +108,9 @@ public EsqlExecutionInfo(Predicate<String> skipUnavailablePredicate, boolean inc

public EsqlExecutionInfo(StreamInput in) throws IOException {
this.overallTook = in.readOptionalTimeValue();
List<EsqlExecutionInfo.Cluster> clusterList = in.readCollectionAsList(EsqlExecutionInfo.Cluster::new);
if (clusterList.isEmpty()) {
this.clusterInfo = ConcurrentCollections.newConcurrentMap();
} else {
Map<String, EsqlExecutionInfo.Cluster> m = ConcurrentCollections.newConcurrentMap();
clusterList.forEach(c -> m.put(c.getClusterAlias(), c));
this.clusterInfo = m;
}
if (in.getTransportVersion().onOrAfter(TransportVersions.V_8_16_0)) {
this.includeCCSMetadata = in.readBoolean();
} else {
this.includeCCSMetadata = false;
}

if (in.getTransportVersion().onOrAfter(TransportVersions.ESQL_RESPONSE_PARTIAL)) {
this.isPartial = in.readBoolean();
} else {
this.isPartial = false;
}

this.clusterInfo = new ConcurrentHashMap<>(in.readMapValues(EsqlExecutionInfo.Cluster::new, Cluster::getClusterAlias));
Copy link
Contributor

@smalyshev smalyshev Jul 24, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The only thing that slightly bothers me here is that we're creating two maps here and there's really no good reason for that, especially given we already have newConcurrentHashMapWithExpectedSize. Maybe this can be improved? Doesn't have to be in this pull, just a thought.

Copy link
Contributor Author

@idegtiarenko idegtiarenko Jul 25, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Previously we created an intermediate list when reading the map, but you are completely right!
We can improve this one by supplying the map factory, updating

this.includeCCSMetadata = in.getTransportVersion().onOrAfter(TransportVersions.V_8_16_0) && in.readBoolean();
Copy link
Contributor

@smalyshev smalyshev Jul 24, 2025 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

To be honest, I prefer the if style here. It's much easier to read and understand and follow the logic. This expression mixes two unrelated logic things (versioning and data deserialization) and while it is true that for booleans it's reduced to the same thing, it creates a mental stumble each time I read this.

this.isPartial = in.getTransportVersion().onOrAfter(TransportVersions.ESQL_RESPONSE_PARTIAL) && in.readBoolean();
this.skipUnavailablePredicate = Predicates.always();
this.relativeStart = null;
if (in.getTransportVersion().onOrAfter(TransportVersions.ESQL_QUERY_PLANNING_DURATION)
Expand All @@ -141,7 +124,7 @@ public EsqlExecutionInfo(StreamInput in) throws IOException {
public void writeTo(StreamOutput out) throws IOException {
out.writeOptionalTimeValue(overallTook);
if (clusterInfo != null) {
out.writeCollection(clusterInfo.values().stream().toList());
out.writeCollection(clusterInfo.values());
Copy link
Contributor Author

@idegtiarenko idegtiarenko Jul 24, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I do not think we need to copy this value before serializing.
clusterInfo is concurrent but it not be changed by the time we serialize it.
Please let me know if you believe otherwise.

Copy link
Contributor

@smalyshev smalyshev Jul 24, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think you are right here, the nodes would only send the info at the end of the computation, AFAIR.

} else {
out.writeCollection(Collections.emptyList());
}
Expand Down