ESQL: Add description to status and profile (elastic#121783) (elastic#121823)

This adds a `task_description` field to `profile` output and task
`status`. This looks like:
```
...
  "profile" : {
    "drivers" : [
      {
        "task_description" : "final",
        "start_millis" : 1738768795349,
        "stop_millis" : 1738768795405,
...
        "task_description" : "node_reduce",
        "start_millis" : 1738768795392,
        "stop_millis" : 1738768795406,
...
        "task_description" : "data",
        "start_millis" : 1738768795391,
        "stop_millis" : 1738768795404,
...
```

Previously you had to look at the signature of the operators in the
driver to figure out what the driver is *doing*. You had to know enough
about how ESQL works to guess. Now you can look at this description to
see what the server *thinks* it is doing. No more manual classification.

This will be useful when debugging failures and performance regressions
because it is much easier to use `jq` to group on it:
```
| jq '.profile[] | group_by(.task_description)[]'
```

Author: nik9000

server/src/main/java/org/elasticsearch/TransportVersions.java

@@ -176,7 +176,7 @@ static TransportVersion def(int id) {
     public static final TransportVersion REMOVE_SNAPSHOT_FAILURES = def(9_002_0_00);
     public static final TransportVersion TRANSPORT_STATS_HANDLING_TIME_REQUIRED = def(9_003_0_00);
     public static final TransportVersion REMOVE_DESIRED_NODE_VERSION = def(9_004_0_00);
-
+    public static final TransportVersion ESQL_DRIVER_TASK_DESCRIPTION = def(9_005_0_00);
     /*
      * STOP! READ THIS FIRST! No, really,
      *        ____ _____ ___  ____  _        ____  _____    _    ____    _____ _   _ ___ ____    _____ ___ ____  ____ _____ _

x-pack/plugin/esql/compute/src/main/java/org/elasticsearch/compute/operator/Driver.java

@@ -52,6 +52,13 @@ public class Driver implements Releasable, Describable {
 
     private final String sessionId;
 
+    /**
+     * Description of the task this driver is running. This description should be
+     * short and meaningful as a grouping identifier. We use the phase of the
+     * query right now: "data", "node_reduce", "final".
+     */
+    private final String taskDescription;
+
     /**
      * The wall clock time when this driver was created in milliseconds since epoch.
      * Compared to {@link #startNanos} this is less accurate and is measured by a
@@ -96,6 +103,10 @@ public class Driver implements Releasable, Describable {
     /**
      * Creates a new driver with a chain of operators.
      * @param sessionId session Id
+     * @param taskDescription Description of the task this driver is running. This
+     *                        description should be short and meaningful as a grouping
+     *                        identifier. We use the phase of the query right now:
+     *                        "data", "node_reduce", "final".
      * @param driverContext the driver context
      * @param source source operator
      * @param intermediateOperators  the chain of operators to execute
@@ -105,6 +116,7 @@ public class Driver implements Releasable, Describable {
      */
     public Driver(
         String sessionId,
+        String taskDescription,
         long startTime,
         long startNanos,
         DriverContext driverContext,
@@ -116,6 +128,7 @@ public Driver(
         Releasable releasable
     ) {
         this.sessionId = sessionId;
+        this.taskDescription = taskDescription;
         this.startTime = startTime;
         this.startNanos = startNanos;
         this.driverContext = driverContext;
@@ -129,6 +142,7 @@ public Driver(
         this.status = new AtomicReference<>(
             new DriverStatus(
                 sessionId,
+                taskDescription,
                 startTime,
                 System.currentTimeMillis(),
                 0,
@@ -150,6 +164,7 @@ public Driver(
      * @param releasable a {@link Releasable} to invoked once the chain of operators has run to completion
      */
     public Driver(
+        String taskDescription,
         DriverContext driverContext,
         SourceOperator source,
         List<Operator> intermediateOperators,
@@ -158,6 +173,7 @@ public Driver(
     ) {
         this(
             "unset",
+            taskDescription,
             System.currentTimeMillis(),
             System.nanoTime(),
             driverContext,
@@ -485,6 +501,7 @@ public DriverProfile profile() {
             throw new IllegalStateException("can only get profile from finished driver");
         }
         return new DriverProfile(
+            status.taskDescription(),
             status.started(),
             status.lastUpdated(),
             finishNanos - startNanos,
@@ -531,6 +548,7 @@ private void updateStatus(long extraCpuNanos, int extraIterations, DriverStatus.
 
             return new DriverStatus(
                 sessionId,
+                taskDescription,
                 startTime,
                 now,
                 prev.cpuNanos() + extraCpuNanos,

x-pack/plugin/esql/compute/src/main/java/org/elasticsearch/compute/operator/DriverProfile.java

@@ -27,6 +27,13 @@
  * Profile results from a single {@link Driver}.
  */
 public class DriverProfile implements Writeable, ChunkedToXContentObject {
+    /**
+     * Description of the task this driver is running. This description should be
+     * short and meaningful as a grouping identifier. We use the phase of the
+     * query right now: "data", "node_reduce", "final".
+     */
+    private final String taskDescription;
+
     /**
      * Millis since epoch when the driver started.
      */
@@ -62,6 +69,7 @@ public class DriverProfile implements Writeable, ChunkedToXContentObject {
     private final DriverSleeps sleeps;
 
     public DriverProfile(
+        String taskDescription,
         long startMillis,
         long stopMillis,
         long tookNanos,
@@ -70,6 +78,7 @@ public DriverProfile(
         List<DriverStatus.OperatorStatus> operators,
         DriverSleeps sleeps
     ) {
+        this.taskDescription = taskDescription;
         this.startMillis = startMillis;
         this.stopMillis = stopMillis;
         this.tookNanos = tookNanos;
@@ -80,6 +89,7 @@ public DriverProfile(
     }
 
     public DriverProfile(StreamInput in) throws IOException {
+        this.taskDescription = in.getTransportVersion().onOrAfter(TransportVersions.ESQL_DRIVER_TASK_DESCRIPTION) ? in.readString() : "";
         if (in.getTransportVersion().onOrAfter(TransportVersions.V_8_16_0)) {
             this.startMillis = in.readVLong();
             this.stopMillis = in.readVLong();
@@ -102,6 +112,9 @@ public DriverProfile(StreamInput in) throws IOException {
 
     @Override
     public void writeTo(StreamOutput out) throws IOException {
+        if (out.getTransportVersion().onOrAfter(TransportVersions.ESQL_DRIVER_TASK_DESCRIPTION)) {
+            out.writeString(taskDescription);
+        }
         if (out.getTransportVersion().onOrAfter(TransportVersions.V_8_16_0)) {
             out.writeVLong(startMillis);
             out.writeVLong(stopMillis);
@@ -115,6 +128,13 @@ public void writeTo(StreamOutput out) throws IOException {
         sleeps.writeTo(out);
     }
 
+    /**
+     * Description of the task this driver is running.
+     */
+    public String taskDescription() {
+        return taskDescription;
+    }
+
     /**
      * Millis since epoch when the driver started.
      */
@@ -169,6 +189,7 @@ public DriverSleeps sleeps() {
     @Override
     public Iterator<? extends ToXContent> toXContentChunked(ToXContent.Params params) {
         return Iterators.concat(ChunkedToXContentHelper.startObject(), Iterators.single((b, p) -> {
+            b.field("task_description", taskDescription);
             b.timestampFieldsFromUnixEpochMillis("start_millis", "start", startMillis);
             b.timestampFieldsFromUnixEpochMillis("stop_millis", "stop", stopMillis);
             b.field("took_nanos", tookNanos);
@@ -197,7 +218,8 @@ public boolean equals(Object o) {
             return false;
         }
         DriverProfile that = (DriverProfile) o;
-        return startMillis == that.startMillis
+        return taskDescription.equals(that.taskDescription)
+            && startMillis == that.startMillis
             && stopMillis == that.stopMillis
             && tookNanos == that.tookNanos
             && cpuNanos == that.cpuNanos
@@ -208,7 +230,7 @@ public boolean equals(Object o) {
 
     @Override
     public int hashCode() {
-        return Objects.hash(startMillis, stopMillis, tookNanos, cpuNanos, iterations, operators, sleeps);
+        return Objects.hash(taskDescription, startMillis, stopMillis, tookNanos, cpuNanos, iterations, operators, sleeps);
     }
 
     @Override

x-pack/plugin/esql/compute/src/main/java/org/elasticsearch/compute/operator/DriverStatus.java

@@ -42,6 +42,11 @@ public class DriverStatus implements Task.Status {
      */
     private final String sessionId;
 
+    /**
+     * Description of the task this driver is running.
+     */
+    private final String taskDescription;
+
     /**
      * Milliseconds since epoch when this driver started.
      */
@@ -83,6 +88,7 @@ public class DriverStatus implements Task.Status {
 
     DriverStatus(
         String sessionId,
+        String taskDescription,
         long started,
         long lastUpdated,
         long cpuTime,
@@ -93,6 +99,7 @@ public class DriverStatus implements Task.Status {
         DriverSleeps sleeps
     ) {
         this.sessionId = sessionId;
+        this.taskDescription = taskDescription;
         this.started = started;
         this.lastUpdated = lastUpdated;
         this.cpuNanos = cpuTime;
@@ -105,6 +112,7 @@ public class DriverStatus implements Task.Status {
 
     public DriverStatus(StreamInput in) throws IOException {
         this.sessionId = in.readString();
+        this.taskDescription = in.getTransportVersion().onOrAfter(TransportVersions.ESQL_DRIVER_TASK_DESCRIPTION) ? in.readString() : "";
         this.started = in.getTransportVersion().onOrAfter(TransportVersions.V_8_14_0) ? in.readLong() : 0;
         this.lastUpdated = in.readLong();
         this.cpuNanos = in.getTransportVersion().onOrAfter(TransportVersions.V_8_14_0) ? in.readVLong() : 0;
@@ -122,6 +130,9 @@ public DriverStatus(StreamInput in) throws IOException {
     @Override
     public void writeTo(StreamOutput out) throws IOException {
         out.writeString(sessionId);
+        if (out.getTransportVersion().onOrAfter(TransportVersions.ESQL_DRIVER_TASK_DESCRIPTION)) {
+            out.writeString(taskDescription);
+        }
         if (out.getTransportVersion().onOrAfter(TransportVersions.V_8_14_0)) {
             out.writeLong(started);
         }
@@ -150,6 +161,15 @@ public String sessionId() {
         return sessionId;
     }
 
+    /**
+     * Description of the task this driver is running. This description should be
+     * short and meaningful as a grouping identifier. We use the phase of the
+     * query right now: "data", "node_reduce", "final".
+     */
+    public String taskDescription() {
+        return taskDescription;
+    }
+
     /**
      * When this {@link Driver} was started.
      */
@@ -211,7 +231,8 @@ public List<OperatorStatus> activeOperators() {
     @Override
     public XContentBuilder toXContent(XContentBuilder builder, Params params) throws IOException {
         builder.startObject();
-        builder.field("sessionId", sessionId);
+        builder.field("session_id", sessionId);
+        builder.field("task_description", taskDescription);
         builder.field("started", DateFieldMapper.DEFAULT_DATE_TIME_FORMATTER.formatMillis(started));
         builder.field("last_updated", DateFieldMapper.DEFAULT_DATE_TIME_FORMATTER.formatMillis(lastUpdated));
         builder.field("cpu_nanos", cpuNanos);
@@ -240,6 +261,7 @@ public boolean equals(Object o) {
         if (o == null || getClass() != o.getClass()) return false;
         DriverStatus that = (DriverStatus) o;
         return sessionId.equals(that.sessionId)
+            && taskDescription.equals(that.taskDescription)
             && started == that.started
             && lastUpdated == that.lastUpdated
             && cpuNanos == that.cpuNanos
@@ -252,7 +274,18 @@ public boolean equals(Object o) {
 
     @Override
     public int hashCode() {
-        return Objects.hash(sessionId, started, lastUpdated, cpuNanos, iterations, status, completedOperators, activeOperators, sleeps);
+        return Objects.hash(
+            sessionId,
+            taskDescription,
+            started,
+            lastUpdated,
+            cpuNanos,
+            iterations,
+            status,
+            completedOperators,
+            activeOperators,
+            sleeps
+        );
     }
 
     @Override

x-pack/plugin/esql/compute/src/test/java/org/elasticsearch/compute/OperatorTests.java

@@ -123,7 +123,7 @@ public void testQueryOperator() throws IOException {
                         }
                     });
                     DriverContext driverContext = driverContext();
-                    drivers.add(new Driver(driverContext, factory.get(driverContext), List.of(), docCollector, () -> {}));
+                    drivers.add(new Driver("test", driverContext, factory.get(driverContext), List.of(), docCollector, () -> {}));
                 }
                 OperatorTestCase.runDriver(drivers);
                 Set<Integer> expectedDocIds = searchForDocIds(reader, query);
@@ -215,6 +215,7 @@ public String toString() {
                     )
                 );
                 Driver driver = new Driver(
+                    "test",
                     driverContext,
                     luceneOperatorFactory(reader, new MatchAllDocsQuery(), LuceneOperator.NO_LIMIT).get(driverContext),
                     operators,
@@ -248,6 +249,7 @@ public void testLimitOperator() {
         DriverContext driverContext = driverContext();
         try (
             var driver = new Driver(
+                "test",
                 driverContext,
                 new SequenceLongBlockSourceOperator(driverContext.blockFactory(), values, 100),
                 List.of((new LimitOperator.Factory(limit)).get(driverContext)),
@@ -335,6 +337,7 @@ public void testHashLookup() {
             var actualPrimeOrds = new ArrayList<>();
             try (
                 var driver = new Driver(
+                    "test",
                     driverContext,
                     new SequenceLongBlockSourceOperator(driverContext.blockFactory(), values, 100),
                     List.of(

x-pack/plugin/esql/compute/src/test/java/org/elasticsearch/compute/aggregation/AggregatorFunctionTestCase.java

@@ -111,6 +111,7 @@ public final void testIgnoresNulls() {
 
         try (
             Driver d = new Driver(
+                "test",
                 driverContext,
                 new NullInsertingSourceOperator(new CannedSourceOperator(input.iterator()), blockFactory),
                 List.of(simple().get(driverContext)),

x-pack/plugin/esql/compute/src/test/java/org/elasticsearch/compute/aggregation/CountDistinctIntAggregatorFunctionTests.java

@@ -65,6 +65,7 @@ public void testRejectsDouble() {
         BlockFactory blockFactory = driverContext.blockFactory();
         try (
             Driver d = new Driver(
+                "test",
                 driverContext,
                 new CannedSourceOperator(Iterators.single(new Page(blockFactory.newDoubleArrayVector(new double[] { 1.0 }, 1).asBlock()))),
                 List.of(simple().get(driverContext)),

x-pack/plugin/esql/compute/src/test/java/org/elasticsearch/compute/aggregation/CountDistinctLongAggregatorFunctionTests.java

@@ -66,6 +66,7 @@ public void testRejectsDouble() {
         BlockFactory blockFactory = driverContext.blockFactory();
         try (
             Driver d = new Driver(
+                "test",
                 driverContext,
                 new CannedSourceOperator(Iterators.single(new Page(blockFactory.newDoubleArrayVector(new double[] { 1.0 }, 1).asBlock()))),
                 List.of(simple().get(driverContext)),

x-pack/plugin/esql/compute/src/test/java/org/elasticsearch/compute/aggregation/SumDoubleAggregatorFunctionTests.java

@@ -53,6 +53,7 @@ public void testOverflowSucceeds() {
         List<Page> results = new ArrayList<>();
         try (
             Driver d = new Driver(
+                "test",
                 driverContext,
                 new SequenceDoubleBlockSourceOperator(driverContext.blockFactory(), DoubleStream.of(Double.MAX_VALUE - 1, 2)),
                 List.of(simple().get(driverContext)),
@@ -71,6 +72,7 @@ public void testSummationAccuracy() {
         List<Page> results = new ArrayList<>();
         try (
             Driver d = new Driver(
+                "test",
                 driverContext,
                 new SequenceDoubleBlockSourceOperator(
                     driverContext.blockFactory(),
@@ -100,6 +102,7 @@ public void testSummationAccuracy() {
         driverContext = driverContext();
         try (
             Driver d = new Driver(
+                "test",
                 driverContext,
                 new SequenceDoubleBlockSourceOperator(driverContext.blockFactory(), DoubleStream.of(values)),
                 List.of(simple().get(driverContext)),
@@ -122,6 +125,7 @@ public void testSummationAccuracy() {
         driverContext = driverContext();
         try (
             Driver d = new Driver(
+                "test",
                 driverContext,
                 new SequenceDoubleBlockSourceOperator(driverContext.blockFactory(), DoubleStream.of(largeValues)),
                 List.of(simple().get(driverContext)),
@@ -141,6 +145,7 @@ public void testSummationAccuracy() {
         driverContext = driverContext();
         try (
             Driver d = new Driver(
+                "test",
                 driverContext,
                 new SequenceDoubleBlockSourceOperator(driverContext.blockFactory(), DoubleStream.of(largeValues)),
                 List.of(simple().get(driverContext)),