github-mirrors/elasticsearch
1
0
Fork 0
mirror of https://github.com/elastic/elasticsearch.git synced 2025-04-24 23:27:25 -04:00
Code Issues Projects Releases Packages Wiki Activity

[ML] deprecate estimated_heap_memory_usage_bytes and replace with model_size_bytes (#80545)

Browse source 
This deprecates estimated_heap_memory_usage_bytes on model put and replaces it with model_size_bytes.

On GET, both fields are returned (unless storing in the index) and are populated with the same field.

For the ml/info API, both fields are returned as well.
This commit is contained in:
Benjamin Trent 2021-11-10 11:24:04 -05:00 • committed by GitHub
parent 7c8a37e3d2
commit 4724553d91
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
22 changed files with 150 additions and 81 deletions
Download patch file Download diff file Expand all files Collapse all files

54
client/rest-high-level/src/main/java/org/elasticsearch/client/ml/inference/TrainedModelConfig.java
View file

@ -42,7 +42,9 @@ public class TrainedModelConfig implements ToXContentObject {
public static final ParseField TAGS = new ParseField("tags");
public static final ParseField METADATA = new ParseField("metadata");
public static final ParseField INPUT = new ParseField("input");
@Deprecated
public static final ParseField ESTIMATED_HEAP_MEMORY_USAGE_BYTES = new ParseField("estimated_heap_memory_usage_bytes");
public static final ParseField MODEL_SIZE_BYTES = new ParseField("model_size_bytes", "estimated_heap_memory_usage_bytes");
public static final ParseField ESTIMATED_OPERATIONS = new ParseField("estimated_operations");
public static final ParseField LICENSE_LEVEL = new ParseField("license_level");
public static final ParseField DEFAULT_FIELD_MAP = new ParseField("default_field_map");
@ -65,7 +67,7 @@ public class TrainedModelConfig implements ToXContentObject {
PARSER.declareStringArray(TrainedModelConfig.Builder::setTags, TAGS);
PARSER.declareObject(TrainedModelConfig.Builder::setMetadata, (p, c) -> p.map(), METADATA);
PARSER.declareObject(TrainedModelConfig.Builder::setInput, (p, c) -> TrainedModelInput.fromXContent(p), INPUT);
PARSER.declareLong(TrainedModelConfig.Builder::setEstimatedHeapMemory, ESTIMATED_HEAP_MEMORY_USAGE_BYTES);
PARSER.declareLong(TrainedModelConfig.Builder::setModelSize, MODEL_SIZE_BYTES);
PARSER.declareLong(TrainedModelConfig.Builder::setEstimatedOperations, ESTIMATED_OPERATIONS);
PARSER.declareString(TrainedModelConfig.Builder::setLicenseLevel, LICENSE_LEVEL);
PARSER.declareObject(TrainedModelConfig.Builder::setDefaultFieldMap, (p, c) -> p.mapStrings(), DEFAULT_FIELD_MAP);
@ -90,7 +92,7 @@ public class TrainedModelConfig implements ToXContentObject {
private final List<String> tags;
private final Map<String, Object> metadata;
private final TrainedModelInput input;
private final Long estimatedHeapMemory;
private final Long modelSize;
private final Long estimatedOperations;
private final String licenseLevel;
private final Map<String, String> defaultFieldMap;
@ -107,7 +109,7 @@ public class TrainedModelConfig implements ToXContentObject {
List<String> tags,
Map<String, Object> metadata,
TrainedModelInput input,
Long estimatedHeapMemory,
Long modelSize,
Long estimatedOperations,
String licenseLevel,
Map<String, String> defaultFieldMap,
@ -123,7 +125,7 @@ public class TrainedModelConfig implements ToXContentObject {
this.tags = tags == null ? null : Collections.unmodifiableList(tags);
this.metadata = metadata == null ? null : Collections.unmodifiableMap(metadata);
this.input = input;
this.estimatedHeapMemory = estimatedHeapMemory;
this.modelSize = modelSize;
this.estimatedOperations = estimatedOperations;
this.licenseLevel = licenseLevel;
this.defaultFieldMap = defaultFieldMap == null ? null : Collections.unmodifiableMap(defaultFieldMap);
@ -170,16 +172,36 @@ public class TrainedModelConfig implements ToXContentObject {
return input;
}
/**
* @deprecated use {@link TrainedModelConfig#getModelSize()} instead
* @return the {@link ByteSizeValue} of the model size if available.
*/
@Deprecated
public ByteSizeValue getEstimatedHeapMemory() {
return estimatedHeapMemory == null ? null : new ByteSizeValue(estimatedHeapMemory);
return modelSize == null ? null : new ByteSizeValue(modelSize);
}
/**
* @deprecated use {@link TrainedModelConfig#getModelSizeBytes()} instead
* @return the model size in bytes if available.
*/
@Deprecated
public Long getEstimatedHeapMemoryBytes() {
return estimatedHeapMemory;
return modelSize;
}
public Long getEstimatedOperations() {
return estimatedOperations;
/**
* @return the {@link ByteSizeValue} of the model size if available.
*/
public ByteSizeValue getModelSize() {
return modelSize == null ? null : new ByteSizeValue(modelSize);
}
/**
* @return the model size in bytes if available.
*/
public Long getModelSizeBytes() {
return modelSize;
}
public String getLicenseLevel() {
@ -228,8 +250,8 @@ public class TrainedModelConfig implements ToXContentObject {
if (input != null) {
builder.field(INPUT.getPreferredName(), input);
}
if (estimatedHeapMemory != null) {
builder.field(ESTIMATED_HEAP_MEMORY_USAGE_BYTES.getPreferredName(), estimatedHeapMemory);
if (modelSize != null) {
builder.field(MODEL_SIZE_BYTES.getPreferredName(), modelSize);
}
if (estimatedOperations != null) {
builder.field(ESTIMATED_OPERATIONS.getPreferredName(), estimatedOperations);
@ -269,7 +291,7 @@ public class TrainedModelConfig implements ToXContentObject {
&& Objects.equals(compressedDefinition, that.compressedDefinition)
&& Objects.equals(tags, that.tags)
&& Objects.equals(input, that.input)
&& Objects.equals(estimatedHeapMemory, that.estimatedHeapMemory)
&& Objects.equals(modelSize, that.modelSize)
&& Objects.equals(estimatedOperations, that.estimatedOperations)
&& Objects.equals(licenseLevel, that.licenseLevel)
&& Objects.equals(defaultFieldMap, that.defaultFieldMap)
@ -288,7 +310,7 @@ public class TrainedModelConfig implements ToXContentObject {
compressedDefinition,
description,
tags,
estimatedHeapMemory,
modelSize,
estimatedOperations,
metadata,
licenseLevel,
@ -310,7 +332,7 @@ public class TrainedModelConfig implements ToXContentObject {
private TrainedModelDefinition definition;
private String compressedDefinition;
private TrainedModelInput input;
private Long estimatedHeapMemory;
private Long modelSize;
private Long estimatedOperations;
private String licenseLevel;
private Map<String, String> defaultFieldMap;
@ -379,8 +401,8 @@ public class TrainedModelConfig implements ToXContentObject {
return this;
}
private Builder setEstimatedHeapMemory(Long estimatedHeapMemory) {
this.estimatedHeapMemory = estimatedHeapMemory;
private Builder setModelSize(Long modelSize) {
this.modelSize = modelSize;
return this;
}
@ -416,7 +438,7 @@ public class TrainedModelConfig implements ToXContentObject {
tags,
metadata,
input,
estimatedHeapMemory,
modelSize,
estimatedOperations,
licenseLevel,
defaultFieldMap,

16
docs/reference/migration/migrate_7_16.asciidoc
View file

@ -13,6 +13,7 @@ See also <<release-highlights>> and <<es-release-notes>>.
* <<breaking_716_tls_changes>>
* <<breaking_716_ilm_changes>>
* <<breaking_716_monitoring_changes>>
* <<breaking_716_api_deprecations>>
* <<breaking_716_settings_deprecations>>
* <<breaking_716_indices_deprecations>>
* <<breaking_716_cluster_deprecations>>
@ -277,6 +278,21 @@ Discontinue the use of the `xpack.monitoring.exporters.*.index.template.create_l
as it will no longer be recognized in the next major release.
====
[discrete]
[[breaking_716_api_deprecations]]
==== REST API deprecations
.The `estimated_heap_memory_usage_bytes` property in the create trained models API is deprecated
[%collapsible]
====
*Details* +
The `estimated_heap_memory_usage_bytes` property in the
{ref}/put-trained-models.html[create trained models API] is deprecated in 7.16.
*Impact* +
Use `model_size_bytes` instead.
====
[discrete]
[[breaking_716_settings_deprecations]]
==== Settings deprecations

13
docs/reference/ml/df-analytics/apis/put-trained-models.asciidoc
View file

@ -48,7 +48,7 @@ include::{es-repo-dir}/ml/ml-shared.asciidoc[tag=model-id]
(Optional, boolean)
If set to `true` and a `compressed_definition` is provided, the request defers
definition decompression and skips relevant validations.
This deferral is useful for systems or users that know a good JVM heap size estimate for their
This deferral is useful for systems or users that know a good byte size estimate for their
model and know that their model is valid and likely won't fail during inference.
@ -373,10 +373,7 @@ An array of `trained_model` objects. Supported trained models are `tree` and
A human-readable description of the {infer} trained model.
`estimated_heap_memory_usage_bytes`::
(Optional, integer)
The estimated heap usage in bytes to keep the trained model in memory. This
property is supported only if `defer_definition_decompression` is `true` or the
model definition is not supplied.
(Optional, integer) deprecated:[7.16.0,Replaced by `model_size_bytes`]
`estimated_operations`::
(Optional, integer)
@ -458,6 +455,12 @@ An array of input field names for the model.
(Optional, object)
An object map that contains metadata about the model.
`model_size_bytes`::
(Optional, integer)
The estimated memory usage in bytes to keep the trained model in memory. This
property is supported only if `defer_definition_decompression` is `true` or the
model definition is not supplied.
`tags`::
(Optional, string)
An array of tags to organize the model.

6
docs/reference/rest-api/usage.asciidoc
View file

@ -163,6 +163,12 @@ GET /_xpack/usage
"prepackaged": 1,
"other": 0
},
"model_size_bytes": {
"min": 0.0,
"max": 0.0,
"avg": 0.0,
"total": 0.0
},
"estimated_heap_memory_usage_bytes": {
"min": 0.0,
"max": 0.0,

6
x-pack/plugin/core/src/main/java/org/elasticsearch/xpack/core/ml/action/PutTrainedModelAction.java
View file

@ -23,7 +23,7 @@ import org.elasticsearch.xpack.core.ml.job.messages.Messages;
import java.io.IOException;
import java.util.Objects;
import static org.elasticsearch.xpack.core.ml.inference.TrainedModelConfig.ESTIMATED_HEAP_MEMORY_USAGE_BYTES;
import static org.elasticsearch.xpack.core.ml.inference.TrainedModelConfig.MODEL_SIZE_BYTES;
public class PutTrainedModelAction extends ActionType<PutTrainedModelAction.Response> {
@ -82,13 +82,13 @@ public class PutTrainedModelAction extends ActionType<PutTrainedModelAction.Resp
@Override
public ActionRequestValidationException validate() {
if (deferDefinitionDecompression && config.getEstimatedHeapMemory() == 0 && config.getCompressedDefinitionIfSet() != null) {
if (deferDefinitionDecompression && config.getModelSize() == 0 && config.getCompressedDefinitionIfSet() != null) {
ActionRequestValidationException validationException = new ActionRequestValidationException();
validationException.addValidationError(
"when ["
+ DEFER_DEFINITION_DECOMPRESSION
+ "] is true and a compressed definition is provided, "
+ ESTIMATED_HEAP_MEMORY_USAGE_BYTES
+ MODEL_SIZE_BYTES
+ " must be set"
);
return validationException;

59
x-pack/plugin/core/src/main/java/org/elasticsearch/xpack/core/ml/inference/TrainedModelConfig.java
View file

@ -61,6 +61,7 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
public static final String MODEL_ALIASES = "model_aliases";
private static final String ESTIMATED_HEAP_MEMORY_USAGE_HUMAN = "estimated_heap_memory_usage";
private static final String MODEL_SIZE_HUMAN = "model_size";
public static final ParseField MODEL_ID = new ParseField("model_id");
public static final ParseField CREATED_BY = new ParseField("created_by");
@ -72,7 +73,12 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
public static final ParseField TAGS = new ParseField("tags");
public static final ParseField METADATA = new ParseField("metadata");
public static final ParseField INPUT = new ParseField("input");
public static final ParseField ESTIMATED_HEAP_MEMORY_USAGE_BYTES = new ParseField("estimated_heap_memory_usage_bytes");
public static final ParseField MODEL_SIZE_BYTES = new ParseField("model_size_bytes");
public static final ParseField MODEL_SIZE_BYTES_WITH_DEPRECATION = new ParseField(
"model_size_bytes",
"estimated_heap_memory_usage_bytes"
);
public static final ParseField DEPRECATED_ESTIMATED_HEAP_MEMORY_USAGE_BYTES = new ParseField("estimated_heap_memory_usage_bytes");
public static final ParseField ESTIMATED_OPERATIONS = new ParseField("estimated_operations");
public static final ParseField LICENSE_LEVEL = new ParseField("license_level");
public static final ParseField DEFAULT_FIELD_MAP = new ParseField("default_field_map");
@ -102,7 +108,14 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
parser.declareObject(TrainedModelConfig.Builder::setMetadata, (p, c) -> p.map(), METADATA);
parser.declareString((trainedModelConfig, s) -> {}, InferenceIndexConstants.DOC_TYPE);
parser.declareObject(TrainedModelConfig.Builder::setInput, (p, c) -> TrainedModelInput.fromXContent(p, ignoreUnknownFields), INPUT);
parser.declareLong(TrainedModelConfig.Builder::setEstimatedHeapMemory, ESTIMATED_HEAP_MEMORY_USAGE_BYTES);
if (ignoreUnknownFields) {
// On reading from the index, we automatically translate to the new field, no need have a deprecation warning
parser.declareLong(TrainedModelConfig.Builder::setModelSize, DEPRECATED_ESTIMATED_HEAP_MEMORY_USAGE_BYTES);
parser.declareLong(TrainedModelConfig.Builder::setModelSize, MODEL_SIZE_BYTES);
} else {
// If this is a new PUT, we should indicate that `estimated_heap_memory_usage_bytes` is deprecated
parser.declareLong(TrainedModelConfig.Builder::setModelSize, MODEL_SIZE_BYTES_WITH_DEPRECATION);
}
parser.declareLong(TrainedModelConfig.Builder::setEstimatedOperations, ESTIMATED_OPERATIONS);
parser.declareObject(
TrainedModelConfig.Builder::setLazyDefinition,
@ -134,7 +147,7 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
private final List<String> tags;
private final Map<String, Object> metadata;
private final TrainedModelInput input;
private final long estimatedHeapMemory;
private final long modelSize;
private final long estimatedOperations;
private final License.OperationMode licenseLevel;
private final Map<String, String> defaultFieldMap;
@ -152,7 +165,7 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
List<String> tags,
Map<String, Object> metadata,
TrainedModelInput input,
Long estimatedHeapMemory,
Long modelSize,
Long estimatedOperations,
String licenseLevel,
Map<String, String> defaultFieldMap,
@ -167,12 +180,10 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
this.tags = Collections.unmodifiableList(ExceptionsHelper.requireNonNull(tags, TAGS));
this.metadata = metadata == null ? null : Collections.unmodifiableMap(metadata);
this.input = ExceptionsHelper.requireNonNull(input, INPUT);
if (ExceptionsHelper.requireNonNull(estimatedHeapMemory, ESTIMATED_HEAP_MEMORY_USAGE_BYTES) < 0) {
throw new IllegalArgumentException(
"[" + ESTIMATED_HEAP_MEMORY_USAGE_BYTES.getPreferredName() + "] must be greater than or equal to 0"
);
if (ExceptionsHelper.requireNonNull(modelSize, MODEL_SIZE_BYTES) < 0) {
throw new IllegalArgumentException("[" + MODEL_SIZE_BYTES.getPreferredName() + "] must be greater than or equal to 0");
}
this.estimatedHeapMemory = estimatedHeapMemory;
this.modelSize = modelSize;
if (ExceptionsHelper.requireNonNull(estimatedOperations, ESTIMATED_OPERATIONS) < 0) {
throw new IllegalArgumentException("[" + ESTIMATED_OPERATIONS.getPreferredName() + "] must be greater than or equal to 0");
}
@ -194,7 +205,7 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
tags = Collections.unmodifiableList(in.readList(StreamInput::readString));
metadata = in.readMap();
input = new TrainedModelInput(in);
estimatedHeapMemory = in.readVLong();
modelSize = in.readVLong();
estimatedOperations = in.readVLong();
licenseLevel = License.OperationMode.parse(in.readString());
if (in.getVersion().onOrAfter(Version.V_7_7_0)) {
@ -299,8 +310,8 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
return new Builder();
}
public long getEstimatedHeapMemory() {
return estimatedHeapMemory;
public long getModelSize() {
return modelSize;
}
public long getEstimatedOperations() {
@ -323,7 +334,7 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
out.writeCollection(tags, StreamOutput::writeString);
out.writeMap(metadata);
input.writeTo(out);
out.writeVLong(estimatedHeapMemory);
out.writeVLong(modelSize);
out.writeVLong(estimatedOperations);
out.writeString(licenseLevel.description());
if (out.getVersion().onOrAfter(Version.V_7_7_0)) {
@ -348,11 +359,15 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
builder.field(CREATED_BY.getPreferredName(), createdBy);
builder.field(VERSION.getPreferredName(), version.toString());
builder.timeField(CREATE_TIME.getPreferredName(), CREATE_TIME.getPreferredName() + "_string", createTime.toEpochMilli());
// If we are NOT storing the model, we should return the deprecated field name
if (params.paramAsBoolean(ToXContentParams.FOR_INTERNAL_STORAGE, false) == false) {
builder.humanReadableField(
ESTIMATED_HEAP_MEMORY_USAGE_BYTES.getPreferredName(),
DEPRECATED_ESTIMATED_HEAP_MEMORY_USAGE_BYTES.getPreferredName(),
ESTIMATED_HEAP_MEMORY_USAGE_HUMAN,
ByteSizeValue.ofBytes(estimatedHeapMemory)
ByteSizeValue.ofBytes(modelSize)
);
}
builder.humanReadableField(MODEL_SIZE_BYTES.getPreferredName(), MODEL_SIZE_HUMAN, ByteSizeValue.ofBytes(modelSize));
builder.field(ESTIMATED_OPERATIONS.getPreferredName(), estimatedOperations);
builder.field(LICENSE_LEVEL.getPreferredName(), licenseLevel.description());
}
@ -403,7 +418,7 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
&& Objects.equals(definition, that.definition)
&& Objects.equals(tags, that.tags)
&& Objects.equals(input, that.input)
&& Objects.equals(estimatedHeapMemory, that.estimatedHeapMemory)
&& Objects.equals(modelSize, that.modelSize)
&& Objects.equals(estimatedOperations, that.estimatedOperations)
&& Objects.equals(licenseLevel, that.licenseLevel)
&& Objects.equals(defaultFieldMap, that.defaultFieldMap)
@ -422,7 +437,7 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
description,
tags,
metadata,
estimatedHeapMemory,
modelSize,
estimatedOperations,
input,
licenseLevel,
@ -441,7 +456,7 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
private List<String> tags = Collections.emptyList();
private Map<String, Object> metadata;
private TrainedModelInput input;
private Long estimatedHeapMemory;
private Long modelSize;
private Long estimatedOperations;
private LazyModelDefinition definition;
private String licenseLevel;
@ -461,7 +476,7 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
this.metadata = config.getMetadata() == null ? null : new HashMap<>(config.getMetadata());
this.input = config.getInput();
this.estimatedOperations = config.estimatedOperations;
this.estimatedHeapMemory = config.estimatedHeapMemory;
this.modelSize = config.modelSize;
this.licenseLevel = config.licenseLevel.description();
this.defaultFieldMap = config.defaultFieldMap == null ? null : new HashMap<>(config.defaultFieldMap);
this.inferenceConfig = config.inferenceConfig;
@ -611,8 +626,8 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
return this;
}
public Builder setEstimatedHeapMemory(long estimatedHeapMemory) {
this.estimatedHeapMemory = estimatedHeapMemory;
public Builder setModelSize(long modelSize) {
this.modelSize = modelSize;
return this;
}
@ -757,7 +772,7 @@ public class TrainedModelConfig implements ToXContentObject, Writeable {
tags,
metadata,
input,
estimatedHeapMemory == null ? 0 : estimatedHeapMemory,
modelSize == null ? 0 : modelSize,
estimatedOperations == null ? 0 : estimatedOperations,
licenseLevel == null ? License.OperationMode.PLATINUM.description() : licenseLevel,
defaultFieldMap,

5
x-pack/plugin/core/src/main/java/org/elasticsearch/xpack/core/ml/inference/persistence/InferenceIndexConstants.java
View file

@ -26,8 +26,11 @@ public final class InferenceIndexConstants {
*
* version: 7.10.0: 000003
* - adds trained_model_metadata object
*
* version: 7.16.0: 000004
* - adds model_size_bytes field as a estimated_heap_memory_usage_bytes replacement
*/
public static final String INDEX_VERSION = "000003";
public static final String INDEX_VERSION = "000004";
public static final String INDEX_NAME_PREFIX = ".ml-inference-";
public static final String INDEX_PATTERN = INDEX_NAME_PREFIX + "*";
public static final String LATEST_INDEX_NAME = INDEX_NAME_PREFIX + INDEX_VERSION;

3
x-pack/plugin/core/src/main/resources/org/elasticsearch/xpack/core/ml/inference_index_mappings.json
View file

@ -38,6 +38,9 @@
"estimated_heap_memory_usage_bytes": {
"type": "long"
},
"model_size_bytes": {
"type": "long"
},
"doc_num": {
"type": "long"
},

2
x-pack/plugin/core/src/test/java/org/elasticsearch/xpack/core/ml/inference/TrainedModelConfigTests.java
View file

@ -63,7 +63,7 @@ public class TrainedModelConfigTests extends AbstractBWCSerializationTestCase<Tr
.setModelId(modelId)
.setCreatedBy(randomAlphaOfLength(10))
.setDescription(randomBoolean() ? null : randomAlphaOfLength(100))
.setEstimatedHeapMemory(randomNonNegativeLong())
.setModelSize(randomNonNegativeLong())
.setEstimatedOperations(randomNonNegativeLong())
.setLicenseLevel(randomFrom(License.OperationMode.PLATINUM.description(), License.OperationMode.BASIC.description()))
.setInferenceConfig(

4
x-pack/plugin/ml/src/internalClusterTest/java/org/elasticsearch/xpack/ml/integration/ChunkedTrainedModelPersisterIT.java
View file

@ -117,7 +117,7 @@ public class ChunkedTrainedModelPersisterIT extends MlSingleNodeTestCase {
TrainedModelConfig storedConfig = getTrainedModelFuture.actionGet();
assertThat(storedConfig.getCompressedDefinition(), equalTo(compressedDefinition));
assertThat(storedConfig.getEstimatedOperations(), equalTo((long) modelSizeInfo.numOperations()));
assertThat(storedConfig.getEstimatedHeapMemory(), equalTo(modelSizeInfo.ramBytesUsed()));
assertThat(storedConfig.getModelSize(), equalTo(modelSizeInfo.ramBytesUsed()));
assertThat(storedConfig.getMetadata(), hasKey("total_feature_importance"));
assertThat(storedConfig.getMetadata(), hasKey("feature_importance_baseline"));
assertThat(storedConfig.getMetadata(), hasKey("hyperparameters"));
@ -141,7 +141,7 @@ public class ChunkedTrainedModelPersisterIT extends MlSingleNodeTestCase {
.setModelId(modelId)
.setVersion(Version.CURRENT)
.setLicenseLevel(License.OperationMode.PLATINUM.description())
.setEstimatedHeapMemory(bytesUsed)
.setModelSize(bytesUsed)
.setEstimatedOperations(operations)
.setInput(TrainedModelInputTests.createRandomInput());
}

8
x-pack/plugin/ml/src/internalClusterTest/java/org/elasticsearch/xpack/ml/integration/ModelInferenceActionIT.java
View file

@ -80,7 +80,7 @@ public class ModelInferenceActionIT extends MlSingleNodeTestCase {
.setLicenseLevel(License.OperationMode.PLATINUM.description())
.setCreateTime(Instant.now())
.setEstimatedOperations(0)
.setEstimatedHeapMemory(0)
.setModelSize(0)
.build();
TrainedModelConfig config2 = buildTrainedModelConfigBuilder(modelId1).setInput(
new TrainedModelInput(Arrays.asList("field.foo", "field.bar", "other.categorical"))
@ -92,7 +92,7 @@ public class ModelInferenceActionIT extends MlSingleNodeTestCase {
)
.setVersion(Version.CURRENT)
.setEstimatedOperations(0)
.setEstimatedHeapMemory(0)
.setModelSize(0)
.setCreateTime(Instant.now())
.build();
AtomicReference<Boolean> putConfigHolder = new AtomicReference<>();
@ -249,7 +249,7 @@ public class ModelInferenceActionIT extends MlSingleNodeTestCase {
.setLicenseLevel(License.OperationMode.PLATINUM.description())
.setCreateTime(Instant.now())
.setEstimatedOperations(0)
.setEstimatedHeapMemory(0)
.setModelSize(0)
.build();
AtomicReference<Boolean> putConfigHolder = new AtomicReference<>();
AtomicReference<Exception> exceptionHolder = new AtomicReference<>();
@ -395,7 +395,7 @@ public class ModelInferenceActionIT extends MlSingleNodeTestCase {
)
.setVersion(Version.CURRENT)
.setEstimatedOperations(0)
.setEstimatedHeapMemory(0)
.setModelSize(0)
.setCreateTime(Instant.now())
.build();
AtomicReference<Boolean> putConfigHolder = new AtomicReference<>();

4
x-pack/plugin/ml/src/internalClusterTest/java/org/elasticsearch/xpack/ml/integration/TrainedModelProviderIT.java
View file

@ -147,7 +147,7 @@ public class TrainedModelProviderIT extends MlSingleNodeTestCase {
.setCreatedBy(config.getCreatedBy())
.setCreateTime(config.getCreateTime())
.setDescription(config.getDescription())
.setEstimatedHeapMemory(config.getEstimatedHeapMemory())
.setModelSize(config.getModelSize())
.setEstimatedOperations(config.getEstimatedOperations())
.setInput(config.getInput())
.setModelId(config.getModelId())
@ -325,7 +325,7 @@ public class TrainedModelProviderIT extends MlSingleNodeTestCase {
.setModelId(modelId)
.setVersion(Version.CURRENT)
.setLicenseLevel(License.OperationMode.PLATINUM.description())
.setEstimatedHeapMemory(0)
.setModelSize(0)
.setEstimatedOperations(0)
.setInput(TrainedModelInputTests.createRandomInput());
}

5
x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/MachineLearningFeatureSet.java
View file

@ -491,7 +491,7 @@ public class MachineLearningFeatureSet implements XPackFeatureSet {
createdByAnalyticsCount++;
}
estimatedOperations.add(trainedModelConfig.getEstimatedOperations());
estimatedMemoryUsageBytes.add(trainedModelConfig.getEstimatedHeapMemory());
estimatedMemoryUsageBytes.add(trainedModelConfig.getModelSize());
}
Map<String, Object> counts = new HashMap<>();
@ -504,9 +504,10 @@ public class MachineLearningFeatureSet implements XPackFeatureSet {
trainedModelsUsage.put("count", counts);
trainedModelsUsage.put(TrainedModelConfig.ESTIMATED_OPERATIONS.getPreferredName(), estimatedOperations.asMap());
trainedModelsUsage.put(
TrainedModelConfig.ESTIMATED_HEAP_MEMORY_USAGE_BYTES.getPreferredName(),
TrainedModelConfig.DEPRECATED_ESTIMATED_HEAP_MEMORY_USAGE_BYTES.getPreferredName(),
estimatedMemoryUsageBytes.asMap()
);
trainedModelsUsage.put(TrainedModelConfig.MODEL_SIZE_BYTES.getPreferredName(), estimatedMemoryUsageBytes.asMap());
inferenceUsage.put("trained_models", trainedModelsUsage);
}

2
x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/action/TransportPutTrainedModelAction.java
View file

@ -159,7 +159,7 @@ public class TransportPutTrainedModelAction extends TransportMasterNodeAction<Re
.setCreatedBy("api_user")
.setLicenseLevel(License.OperationMode.PLATINUM.description());
if (hasModelDefinition) {
trainedModelConfigBuilder.setEstimatedHeapMemory(request.getTrainedModelConfig().getModelDefinition().ramBytesUsed())
trainedModelConfigBuilder.setModelSize(request.getTrainedModelConfig().getModelDefinition().ramBytesUsed())
.setEstimatedOperations(request.getTrainedModelConfig().getModelDefinition().getTrainedModel().estimatedNumOperations());
}
TrainedModelConfig trainedModelConfig = trainedModelConfigBuilder.build();

2
x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/dataframe/process/ChunkedTrainedModelPersister.java
View file

@ -315,7 +315,7 @@ public class ChunkedTrainedModelPersister {
XContentHelper.convertToMap(JsonXContent.jsonXContent, analytics.toString(), true)
)
)
.setEstimatedHeapMemory(modelSize.ramBytesUsed() + customProcessorSize)
.setModelSize(modelSize.ramBytesUsed() + customProcessorSize)
.setEstimatedOperations(modelSize.numOperations())
.setInput(new TrainedModelInput(fieldNamesWithoutDependentVariable))
.setLicenseLevel(License.OperationMode.PLATINUM.description())

12
x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/inference/loadingservice/ModelLoadingService.java
View file

@ -323,7 +323,7 @@ public class ModelLoadingService implements ClusterStateListener {
private void loadModel(String modelId, Consumer consumer) {
provider.getTrainedModel(modelId, GetTrainedModelsAction.Includes.empty(), ActionListener.wrap(trainedModelConfig -> {
trainedModelCircuitBreaker.addEstimateBytesAndMaybeBreak(trainedModelConfig.getEstimatedHeapMemory(), modelId);
trainedModelCircuitBreaker.addEstimateBytesAndMaybeBreak(trainedModelConfig.getModelSize(), modelId);
provider.getTrainedModelForInference(modelId, consumer == Consumer.INTERNAL, ActionListener.wrap(inferenceDefinition -> {
try {
// Since we have used the previously stored estimate to help guard against OOM we need
@ -338,7 +338,7 @@ public class ModelLoadingService implements ClusterStateListener {
handleLoadSuccess(modelId, consumer, trainedModelConfig, inferenceDefinition);
}, failure -> {
// We failed to get the definition, remove the initial estimation.
trainedModelCircuitBreaker.addWithoutBreaking(-trainedModelConfig.getEstimatedHeapMemory());
trainedModelCircuitBreaker.addWithoutBreaking(-trainedModelConfig.getModelSize());
logger.warn(new ParameterizedMessage("[{}] failed to load model definition", modelId), failure);
handleLoadFailure(modelId, failure);
}));
@ -353,7 +353,7 @@ public class ModelLoadingService implements ClusterStateListener {
// by a simulated pipeline
provider.getTrainedModel(modelId, GetTrainedModelsAction.Includes.empty(), ActionListener.wrap(trainedModelConfig -> {
// Verify we can pull the model into memory without causing OOM
trainedModelCircuitBreaker.addEstimateBytesAndMaybeBreak(trainedModelConfig.getEstimatedHeapMemory(), modelId);
trainedModelCircuitBreaker.addEstimateBytesAndMaybeBreak(trainedModelConfig.getModelSize(), modelId);
provider.getTrainedModelForInference(modelId, consumer == Consumer.INTERNAL, ActionListener.wrap(inferenceDefinition -> {
InferenceConfig inferenceConfig = trainedModelConfig.getInferenceConfig() == null
? inferenceConfigFromTargetType(inferenceDefinition.getTargetType())
@ -381,7 +381,7 @@ public class ModelLoadingService implements ClusterStateListener {
},
// Failure getting the definition, remove the initial estimation value
e -> {
trainedModelCircuitBreaker.addWithoutBreaking(-trainedModelConfig.getEstimatedHeapMemory());
trainedModelCircuitBreaker.addWithoutBreaking(-trainedModelConfig.getModelSize());
modelActionListener.onFailure(e);
}
));
@ -393,14 +393,14 @@ public class ModelLoadingService implements ClusterStateListener {
InferenceDefinition inferenceDefinition,
TrainedModelConfig trainedModelConfig
) throws CircuitBreakingException {
long estimateDiff = inferenceDefinition.ramBytesUsed() - trainedModelConfig.getEstimatedHeapMemory();
long estimateDiff = inferenceDefinition.ramBytesUsed() - trainedModelConfig.getModelSize();
if (estimateDiff < 0) {
trainedModelCircuitBreaker.addWithoutBreaking(estimateDiff);
} else if (estimateDiff > 0) { // rare case where estimate is now HIGHER
try {
trainedModelCircuitBreaker.addEstimateBytesAndMaybeBreak(estimateDiff, modelId);
} catch (CircuitBreakingException ex) { // if we failed here, we should remove the initial estimate as well
trainedModelCircuitBreaker.addWithoutBreaking(-trainedModelConfig.getEstimatedHeapMemory());
trainedModelCircuitBreaker.addWithoutBreaking(-trainedModelConfig.getModelSize());
throw ex;
}
}

2
x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/rest/cat/RestCatTrainedModelsAction.java
View file

@ -265,7 +265,7 @@ public class RestCatTrainedModelsAction extends AbstractCatAction {
// Trained Model Info
table.addCell(config.getModelId());
table.addCell(config.getCreatedBy());
table.addCell(ByteSizeValue.ofBytes(config.getEstimatedHeapMemory()));
table.addCell(ByteSizeValue.ofBytes(config.getModelSize()));
table.addCell(config.getEstimatedOperations());
table.addCell(config.getLicenseLevel());
table.addCell(config.getCreateTime());

2
x-pack/plugin/ml/src/main/resources/org/elasticsearch/xpack/ml/inference/persistence/lang_ident_model_1.json
View file

@ -15,7 +15,7 @@
]
},
"inference_config" : {"classification" : {}},
"estimated_heap_memory_usage_bytes" : 1053992,
"model_size_bytes" : 1053992,
"estimated_operations" : 39629,
"license_level" : "basic"
}

8
x-pack/plugin/ml/src/test/java/org/elasticsearch/xpack/ml/MachineLearningFeatureSetTests.java
View file

@ -293,22 +293,22 @@ public class MachineLearningFeatureSetTests extends ESTestCase {
);
TrainedModelConfig trainedModel1 = TrainedModelConfigTests.createTestInstance("model_1")
.setEstimatedHeapMemory(100)
.setModelSize(100)
.setEstimatedOperations(200)
.setMetadata(Collections.singletonMap("analytics_config", "anything"))
.build();
TrainedModelConfig trainedModel2 = TrainedModelConfigTests.createTestInstance("model_2")
.setEstimatedHeapMemory(200)
.setModelSize(200)
.setEstimatedOperations(400)
.setMetadata(Collections.singletonMap("analytics_config", "anything"))
.build();
TrainedModelConfig trainedModel3 = TrainedModelConfigTests.createTestInstance("model_3")
.setEstimatedHeapMemory(300)
.setModelSize(300)
.setEstimatedOperations(600)
.build();
TrainedModelConfig trainedModel4 = TrainedModelConfigTests.createTestInstance("model_4")
.setTags(Collections.singletonList("prepackaged"))
.setEstimatedHeapMemory(1000)
.setModelSize(1000)
.setEstimatedOperations(2000)
.build();
givenTrainedModels(Arrays.asList(trainedModel1, trainedModel2, trainedModel3, trainedModel4));

2
x-pack/plugin/ml/src/test/java/org/elasticsearch/xpack/ml/dataframe/process/ChunkedTrainedModelPersisterTests.java
View file

@ -138,7 +138,7 @@ public class ChunkedTrainedModelPersisterTests extends ESTestCase {
assertThat(storedModel.getTags(), contains(JOB_ID));
assertThat(storedModel.getDescription(), equalTo(JOB_DESCRIPTION));
assertThat(storedModel.getModelDefinition(), is(nullValue()));
assertThat(storedModel.getEstimatedHeapMemory(), equalTo(modelSizeInfo.ramBytesUsed()));
assertThat(storedModel.getModelSize(), equalTo(modelSizeInfo.ramBytesUsed()));
assertThat(storedModel.getEstimatedOperations(), equalTo((long) modelSizeInfo.numOperations()));
if (analyticsConfig.getAnalysis() instanceof Classification) {
assertThat(storedModel.getInferenceConfig().getName(), equalTo("classification"));

4
x-pack/plugin/ml/src/test/java/org/elasticsearch/xpack/ml/inference/loadingservice/ModelLoadingServiceTests.java
View file

@ -658,7 +658,7 @@ public class ModelLoadingServiceTests extends ESTestCase {
when(trainedModelConfig.getModelId()).thenReturn(modelId);
when(trainedModelConfig.getInferenceConfig()).thenReturn(ClassificationConfig.EMPTY_PARAMS);
when(trainedModelConfig.getInput()).thenReturn(new TrainedModelInput(Arrays.asList("foo", "bar", "baz")));
when(trainedModelConfig.getEstimatedHeapMemory()).thenReturn(size);
when(trainedModelConfig.getModelSize()).thenReturn(size);
doAnswer(invocationOnMock -> {
@SuppressWarnings("rawtypes")
ActionListener listener = (ActionListener) invocationOnMock.getArguments()[2];
@ -684,7 +684,7 @@ public class ModelLoadingServiceTests extends ESTestCase {
}).when(trainedModelProvider).getTrainedModel(eq(modelId), eq(GetTrainedModelsAction.Includes.empty()), any());
} else {
TrainedModelConfig trainedModelConfig = mock(TrainedModelConfig.class);
when(trainedModelConfig.getEstimatedHeapMemory()).thenReturn(0L);
when(trainedModelConfig.getModelSize()).thenReturn(0L);
doAnswer(invocationOnMock -> {
@SuppressWarnings("rawtypes")
ActionListener listener = (ActionListener) invocationOnMock.getArguments()[2];

4
x-pack/plugin/src/yamlRestTest/resources/rest-api-spec/test/ml/inference_crud.yml
View file

@ -965,14 +965,14 @@ setup:
"description": "model for tests",
"input": {"field_names": ["field1", "field2"]},
"inference_config": {"classification": {}},
"estimated_heap_memory_usage_bytes": 1024,
"model_size_bytes": 1024,
"compressed_definition": "H4sIAAAAAAAAAEy92a5mW26l9y55HWdj9o3u9RS+SMil4yrBUgpIpywY9fLmR3LMFSpI"
}
---
"Test put with defer_definition_decompression with invalid compression definition and no memory estimate":
- do:
catch: /when \[defer_definition_decompression\] is true and a compressed definition is provided, estimated_heap_memory_usage_bytes must be set/
catch: /when \[defer_definition_decompression\] is true and a compressed definition is provided, model_size_bytes must be set/
ml.put_trained_model:
defer_definition_decompression: true
model_id: my-regression-model-compressed-failed