Skip to content
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

[ML] add _cat/ml/trained_models API #51529

Merged
Merged
Show file tree
Hide file tree
Changes from 5 commits
Commits
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
Original file line number Diff line number Diff line change
Expand Up @@ -103,6 +103,14 @@ public String getModelId() {
return modelId;
}

public IngestStats getIngestStats() {
return ingestStats;
}

public int getPipelineCount() {
return pipelineCount;
}

@Override
public XContentBuilder toXContent(XContentBuilder builder, Params params) throws IOException {
builder.startObject();
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -259,6 +259,7 @@
import org.elasticsearch.xpack.ml.rest.calendar.RestPutCalendarAction;
import org.elasticsearch.xpack.ml.rest.calendar.RestPutCalendarJobAction;
import org.elasticsearch.xpack.ml.rest.cat.RestCatJobsAction;
import org.elasticsearch.xpack.ml.rest.cat.RestCatTrainedModelsAction;
import org.elasticsearch.xpack.ml.rest.datafeeds.RestDeleteDatafeedAction;
import org.elasticsearch.xpack.ml.rest.datafeeds.RestGetDatafeedStatsAction;
import org.elasticsearch.xpack.ml.rest.datafeeds.RestGetDatafeedsAction;
Expand Down Expand Up @@ -773,7 +774,8 @@ public List<RestHandler> getRestHandlers(Settings settings, RestController restC
new RestGetTrainedModelsStatsAction(restController),
new RestPutTrainedModelAction(restController),
// CAT Handlers
new RestCatJobsAction(restController)
new RestCatJobsAction(restController),
new RestCatTrainedModelsAction(restController)
);
}

Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -177,6 +177,7 @@ private TrainedModelConfig createTrainedModelConfig(TrainedModelDefinition.Build
.setCreatedBy(XPackUser.NAME)
.setVersion(Version.CURRENT)
.setCreateTime(createTime)
// NOTE: GET _cat/ml/trained_models relies on the creating analytics ID being in the tags
.setTags(Collections.singletonList(analytics.getId()))
.setDescription(analytics.getDescription())
.setMetadata(Collections.singletonMap("analytics_config",
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1,284 @@
/*
* Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
* or more contributor license agreements. Licensed under the Elastic License;
* you may not use this file except in compliance with the Elastic License.
*/
package org.elasticsearch.xpack.ml.rest.cat;

import org.elasticsearch.action.ActionListener;
import org.elasticsearch.action.ActionResponse;
import org.elasticsearch.action.support.GroupedActionListener;
import org.elasticsearch.client.node.NodeClient;
import org.elasticsearch.cluster.metadata.MetaData;
import org.elasticsearch.common.Strings;
import org.elasticsearch.common.Table;
import org.elasticsearch.common.unit.ByteSizeValue;
import org.elasticsearch.common.unit.TimeValue;
import org.elasticsearch.rest.RestController;
import org.elasticsearch.rest.RestRequest;
import org.elasticsearch.rest.RestResponse;
import org.elasticsearch.rest.action.RestResponseListener;
import org.elasticsearch.rest.action.cat.AbstractCatAction;
import org.elasticsearch.rest.action.cat.RestTable;
import org.elasticsearch.xpack.core.action.util.PageParams;
import org.elasticsearch.xpack.core.ml.action.GetDataFrameAnalyticsAction;
import org.elasticsearch.xpack.core.ml.action.GetTrainedModelsAction;
import org.elasticsearch.xpack.core.ml.action.GetTrainedModelsStatsAction;
import org.elasticsearch.xpack.core.ml.dataframe.DataFrameAnalyticsConfig;
import org.elasticsearch.xpack.core.ml.dataframe.analyses.DataFrameAnalysis;
import org.elasticsearch.xpack.core.ml.inference.TrainedModelConfig;
import org.elasticsearch.xpack.core.security.user.XPackUser;

import java.util.Collection;
import java.util.Collections;
import java.util.HashSet;
import java.util.List;
import java.util.Map;
import java.util.Set;
import java.util.function.Function;
import java.util.stream.Collectors;

import static org.elasticsearch.rest.RestRequest.Method.GET;
import static org.elasticsearch.xpack.core.ml.action.GetTrainedModelsAction.Request.ALLOW_NO_MATCH;

public class RestCatTrainedModelsAction extends AbstractCatAction {

public RestCatTrainedModelsAction(RestController controller) {
controller.registerHandler(GET, "_cat/ml/trained_models/{" + TrainedModelConfig.MODEL_ID.getPreferredName() + "}", this);
controller.registerHandler(GET, "_cat/ml/trained_models", this);
}

@Override
public String getName() {
return "cat_ml_get_trained_models_action";
}

@Override
protected RestChannelConsumer doCatRequest(RestRequest restRequest, NodeClient client) {
String modelId = restRequest.param(TrainedModelConfig.MODEL_ID.getPreferredName());
if (Strings.isNullOrEmpty(modelId)) {
modelId = MetaData.ALL;
}
GetTrainedModelsStatsAction.Request statsRequest = new GetTrainedModelsStatsAction.Request(modelId);
GetTrainedModelsAction.Request modelsAction = new GetTrainedModelsAction.Request(modelId, false, null);
if (restRequest.hasParam(PageParams.FROM.getPreferredName()) || restRequest.hasParam(PageParams.SIZE.getPreferredName())) {
Copy link
Member

Choose a reason for hiding this comment

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

This is the only cat action that supports paging

Copy link
Member Author

Choose a reason for hiding this comment

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

Correct, and it probably should as there is a limit of 10K when reading configs from an index.
Our _cat APIs are the only ones that return data that are stored in indices (i think)

statsRequest.setPageParams(new PageParams(restRequest.paramAsInt(PageParams.FROM.getPreferredName(), PageParams.DEFAULT_FROM),
restRequest.paramAsInt(PageParams.SIZE.getPreferredName(), PageParams.DEFAULT_SIZE)));
modelsAction.setPageParams(new PageParams(restRequest.paramAsInt(PageParams.FROM.getPreferredName(), PageParams.DEFAULT_FROM),
restRequest.paramAsInt(PageParams.SIZE.getPreferredName(), PageParams.DEFAULT_SIZE)));
}
statsRequest.setAllowNoResources(true);
modelsAction.setAllowNoResources(restRequest.paramAsBoolean(ALLOW_NO_MATCH.getPreferredName(),
statsRequest.isAllowNoResources()));

return channel -> {
final ActionListener<Table> listener = ActionListener.notifyOnce(new RestResponseListener<>(channel) {
@Override
public RestResponse buildResponse(final Table table) throws Exception {
return RestTable.buildResponse(table, channel);
}
});

client.execute(GetTrainedModelsAction.INSTANCE, modelsAction, ActionListener.wrap(
trainedModels -> {
final List<TrainedModelConfig> trainedModelConfigs = trainedModels.getResources().results();

Set<String> potentialAnalyticsIds = new HashSet<>();
// Analytics Configs are created by the XPackUser
trainedModelConfigs.stream()
.filter(c -> XPackUser.NAME.equals(c.getCreatedBy()))
Copy link
Member

Choose a reason for hiding this comment

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

I think we want a better way of differentiating user models and DFA models in the future. Maybe a reserved tag for DFA models

Copy link
Member Author

Choose a reason for hiding this comment

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

@davidkyle possibly. Users cannot set the created_by, and XPackUser.NAME is a reserved name.

But I see your point for models we provide as a resource. Those weren't created by a DFA.

.forEach(c -> potentialAnalyticsIds.addAll(c.getTags()));


// Find the related DataFrameAnalyticsConfigs
String requestIdPattern = Strings.collectionToDelimitedString(potentialAnalyticsIds, "*,") + "*";

final GroupedActionListener<ActionResponse> groupedListener = createGroupedListener(restRequest,
2,
trainedModels.getResources().results(),
listener);

client.execute(GetTrainedModelsStatsAction.INSTANCE,
statsRequest,
ActionListener.wrap(groupedListener::onResponse, groupedListener::onFailure));
Copy link
Member

Choose a reason for hiding this comment

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

Do you need the wrap?

Copy link
Member Author

Choose a reason for hiding this comment

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

Yes, generic types get upset if there is no wrapper.


GetDataFrameAnalyticsAction.Request dataFrameAnalyticsRequest =
new GetDataFrameAnalyticsAction.Request(requestIdPattern);
dataFrameAnalyticsRequest.setAllowNoResources(true);
dataFrameAnalyticsRequest.setPageParams(new PageParams(0, potentialAnalyticsIds.size()));
client.execute(GetDataFrameAnalyticsAction.INSTANCE,
dataFrameAnalyticsRequest,
ActionListener.wrap(groupedListener::onResponse, groupedListener::onFailure));
},
listener::onFailure
));
};
}

@Override
protected void documentation(StringBuilder sb) {
sb.append("/_cat/ml/trained_models\n");
sb.append("/_cat/ml/trained_models/{model_id}\n");
}

@Override
protected Table getTableWithHeader(RestRequest request) {
Table table = new Table();
table.startHeaders();

// Trained Model Info
table.addCell("id", TableColumnAttributeBuilder.builder().setDescription("the trained model id").build());
table.addCell("created_by", TableColumnAttributeBuilder.builder("who created the model", false)
.setAliases("c", "createdBy")
.setTextAlignment(TableColumnAttributeBuilder.TextAlign.RIGHT)
.build());
table.addCell("heap_size", TableColumnAttributeBuilder.builder()
.setDescription("the estimated heap size to keep the model in memory")
.setAliases("hs","modelHeapSize")
.build());
table.addCell("operations", TableColumnAttributeBuilder.builder()
.setDescription("the estimated number of operations to use the model")
.setAliases("o", "modelOperations")
.build());
table.addCell("license", TableColumnAttributeBuilder.builder("The license level of the model", false)
.setAliases("l")
.build());
table.addCell("create_time", TableColumnAttributeBuilder.builder("The time the model was created", false)
.setAliases("ct")
.build());
table.addCell("version", TableColumnAttributeBuilder.builder("The version of Elasticsearch when the model was created", false)
.setAliases("v")
.build());
table.addCell("description", TableColumnAttributeBuilder.builder("The model description", false)
.setAliases("d")
.build());

// Trained Model Stats
table.addCell("ingest.pipelines", TableColumnAttributeBuilder.builder("The number of pipelines referencing the model")
.setAliases("ip", "ingestPipelines")
.build());
table.addCell("ingest.count", TableColumnAttributeBuilder.builder("The total number of docs processed by the model", false)
.setAliases("ic", "ingestCount")
.build());
table.addCell("ingest.time", TableColumnAttributeBuilder.builder(
"The total time spent processing docs with this model",
false)
.setAliases("it", "ingestTime")
.build());
table.addCell("ingest.current", TableColumnAttributeBuilder.builder(
"The total documents currently being handled by the model",
false)
.setAliases("icurr", "ingestCurrent")
.build());
table.addCell("ingest.failed", TableColumnAttributeBuilder.builder(
"The total count of failed ingest attempts with this model",
false)
.setAliases("if", "ingestFailed")
.build());

table.addCell("data_frame.id", TableColumnAttributeBuilder.builder(
"The data frame analytics config id that created the model (if still available)",
false)
.setAliases("dfid", "dataFrameAnalytics")
.build());
table.addCell("data_frame.create_time", TableColumnAttributeBuilder.builder(
"The time the data frame analytics config was created",
false)
.setAliases("dft", "dataFrameAnalyticsTime")
.build());
table.addCell("data_frame.source_index", TableColumnAttributeBuilder.builder(
"The source index used to train in the data frame analysis",
false)
.setAliases("dfsi", "dataFrameAnalyticsSrcIndex")
.build());
table.addCell("data_frame.analysis", TableColumnAttributeBuilder.builder(
"The analysis used by the data frame to build the model",
false)
.setAliases("dfa", "dataFrameAnalyticsAnalysis")
.build());

table.endHeaders();
return table;
}

private GroupedActionListener<ActionResponse> createGroupedListener(final RestRequest request,
final int size,
final List<TrainedModelConfig> configs,
final ActionListener<Table> listener) {
return new GroupedActionListener<>(new ActionListener<>() {
@Override
public void onResponse(final Collection<ActionResponse> responses) {
GetTrainedModelsStatsAction.Response statsResponse = extractResponse(responses, GetTrainedModelsStatsAction.Response.class);
GetDataFrameAnalyticsAction.Response analytics = extractResponse(responses, GetDataFrameAnalyticsAction.Response.class);
listener.onResponse(buildTable(request,
statsResponse.getResources().results(),
configs,
analytics == null ? Collections.emptyList() : analytics.getResources().results()));
}

@Override
public void onFailure(final Exception e) {
listener.onFailure(e);
}
}, size);
}


private Table buildTable(RestRequest request,
List<GetTrainedModelsStatsAction.Response.TrainedModelStats> stats,
List<TrainedModelConfig> configs,
List<DataFrameAnalyticsConfig> analyticsConfigs) {
Table table = getTableWithHeader(request);
assert configs.size() == stats.size();

Map<String, DataFrameAnalyticsConfig> analyticsMap = analyticsConfigs.stream()
.collect(Collectors.toMap(DataFrameAnalyticsConfig::getId, Function.identity()));
Map<String, GetTrainedModelsStatsAction.Response.TrainedModelStats> statsMap = stats.stream()
.collect(Collectors.toMap(GetTrainedModelsStatsAction.Response.TrainedModelStats::getModelId, Function.identity()));

configs.forEach(config -> {
table.startRow();
// Trained Model Info
table.addCell(config.getModelId());
table.addCell(config.getCreatedBy());
table.addCell(new ByteSizeValue(config.getEstimatedHeapMemory()));
table.addCell(config.getEstimatedOperations());
table.addCell(config.getLicenseLevel());
table.addCell(config.getCreateTime());
table.addCell(config.getVersion().toString());
table.addCell(config.getDescription());

GetTrainedModelsStatsAction.Response.TrainedModelStats modelStats = statsMap.get(config.getModelId());
table.addCell(modelStats.getPipelineCount());
boolean hasIngestStats = modelStats != null && modelStats.getIngestStats() != null;
table.addCell(hasIngestStats ? modelStats.getIngestStats().getTotalStats().getIngestCount() : 0);
table.addCell(hasIngestStats ?
TimeValue.timeValueMillis(modelStats.getIngestStats().getTotalStats().getIngestTimeInMillis()) :
TimeValue.timeValueMillis(0));
table.addCell(hasIngestStats ? modelStats.getIngestStats().getTotalStats().getIngestCurrent() : 0);
table.addCell(hasIngestStats ? modelStats.getIngestStats().getTotalStats().getIngestFailedCount() : 0);

DataFrameAnalyticsConfig dataFrameAnalyticsConfig = config.getTags()
.stream()
.filter(analyticsMap::containsKey)
.map(analyticsMap::get)
.findFirst()
.orElse(null);
table.addCell(dataFrameAnalyticsConfig == null ? "__none__" : dataFrameAnalyticsConfig.getId());
table.addCell(dataFrameAnalyticsConfig == null ? null : dataFrameAnalyticsConfig.getCreateTime());
table.addCell(dataFrameAnalyticsConfig == null ?
null :
Strings.arrayToCommaDelimitedString(dataFrameAnalyticsConfig.getSource().getIndex()));
DataFrameAnalysis analysis = dataFrameAnalyticsConfig == null ? null : dataFrameAnalyticsConfig.getAnalysis();
table.addCell(analysis == null ? null : analysis.getWriteableName());

table.endRow();
});
return table;
}

@SuppressWarnings("unchecked")
private static <A extends ActionResponse> A extractResponse(final Collection<? extends ActionResponse> responses, Class<A> c) {
return (A) responses.stream().filter(c::isInstance).findFirst().get();
}
}
Loading