Search in sources :

Example 1 with ManagedExecution

use of com.bakdata.conquery.models.execution.ManagedExecution in project conquery by bakdata.

the class QueryCleanupTask method execute.

@Override
public void execute(Map<String, List<String>> parameters, PrintWriter output) throws Exception {
    Duration queryExpiration = this.queryExpiration;
    if (parameters.containsKey(EXPIRATION_PARAM)) {
        if (parameters.get(EXPIRATION_PARAM).size() > 1) {
            log.warn("Will not respect more than one expiration time. Have `{}`", parameters.get(EXPIRATION_PARAM));
        }
        queryExpiration = Duration.parse(parameters.get(EXPIRATION_PARAM).get(0));
    }
    if (queryExpiration == null) {
        throw new IllegalArgumentException("Query Expiration may not be null");
    }
    log.info("Starting deletion of queries older than {} of {}", queryExpiration, storage.getAllExecutions().size());
    // Iterate for as long as no changes are needed (this is because queries can be referenced by other queries)
    while (true) {
        final QueryUtils.AllReusedFinder reusedChecker = new QueryUtils.AllReusedFinder();
        Set<ManagedExecution<?>> toDelete = new HashSet<>();
        for (ManagedExecution<?> execution : storage.getAllExecutions()) {
            // Gather all referenced queries via reused checker.
            if (execution instanceof ManagedQuery) {
                ((ManagedQuery) execution).getQuery().visit(reusedChecker);
            } else if (execution instanceof ManagedForm) {
                ((ManagedForm) execution).getFlatSubQueries().values().forEach(q -> q.getQuery().visit(reusedChecker));
            }
            if (execution.isShared()) {
                continue;
            }
            log.trace("{} is not shared", execution.getId());
            if (ArrayUtils.isNotEmpty(execution.getTags())) {
                continue;
            }
            log.trace("{} has no tags", execution.getId());
            if (execution.getLabel() != null && !isDefaultLabel(execution.getLabel())) {
                continue;
            }
            log.trace("{} has no label", execution.getId());
            if (LocalDateTime.now().minus(queryExpiration).isBefore(execution.getCreationTime())) {
                continue;
            }
            log.trace("{} is not older than {}.", execution.getId(), queryExpiration);
            toDelete.add(execution);
        }
        // remove all queries referenced in reused queries.
        final Collection<ManagedExecution<?>> referenced = reusedChecker.getReusedElements().stream().map(CQReusedQuery::getQueryId).map(storage::getExecution).collect(Collectors.toSet());
        toDelete.removeAll(referenced);
        if (toDelete.isEmpty()) {
            log.info("No queries to delete");
            break;
        }
        log.info("Deleting {} Executions", toDelete.size());
        for (ManagedExecution<?> execution : toDelete) {
            log.trace("Deleting Execution[{}]", execution.getId());
            storage.removeExecution(execution.getId());
        }
    }
}
Also used : PrintWriter(java.io.PrintWriter) ManagedQuery(com.bakdata.conquery.models.query.ManagedQuery) ManagedForm(com.bakdata.conquery.models.forms.managed.ManagedForm) Predicate(java.util.function.Predicate) Collection(java.util.Collection) LocalDateTime(java.time.LocalDateTime) Set(java.util.Set) ArrayUtils(org.apache.commons.lang3.ArrayUtils) Collectors(java.util.stream.Collectors) HashSet(java.util.HashSet) List(java.util.List) Slf4j(lombok.extern.slf4j.Slf4j) ManagedExecution(com.bakdata.conquery.models.execution.ManagedExecution) Task(io.dropwizard.servlets.tasks.Task) Duration(java.time.Duration) Map(java.util.Map) QueryUtils(com.bakdata.conquery.util.QueryUtils) CQReusedQuery(com.bakdata.conquery.apiv1.query.concept.specific.CQReusedQuery) Pattern(java.util.regex.Pattern) MetaStorage(com.bakdata.conquery.io.storage.MetaStorage) CQReusedQuery(com.bakdata.conquery.apiv1.query.concept.specific.CQReusedQuery) Duration(java.time.Duration) QueryUtils(com.bakdata.conquery.util.QueryUtils) ManagedForm(com.bakdata.conquery.models.forms.managed.ManagedForm) ManagedQuery(com.bakdata.conquery.models.query.ManagedQuery) ManagedExecution(com.bakdata.conquery.models.execution.ManagedExecution) HashSet(java.util.HashSet)

Example 2 with ManagedExecution

use of com.bakdata.conquery.models.execution.ManagedExecution in project conquery by bakdata.

the class QueryProcessor method postQuery.

/**
 * Creates a query for all datasets, then submits it for execution on the
 * intended dataset.
 */
public ManagedExecution<?> postQuery(Dataset dataset, QueryDescription query, Subject subject) {
    log.info("Query posted on Dataset[{}] by User[{{}].", dataset.getId(), subject.getId());
    // This maps works as long as we have query visitors that are not configured in anyway.
    // So adding a visitor twice would replace the previous one but both would have yielded the same result.
    // For the future a better data structure might be desired that also regards similar QueryVisitors of different configuration
    ClassToInstanceMap<QueryVisitor> visitors = MutableClassToInstanceMap.create();
    query.addVisitors(visitors);
    // Initialize checks that need to traverse the query tree
    visitors.putInstance(QueryUtils.OnlyReusingChecker.class, new QueryUtils.OnlyReusingChecker());
    visitors.putInstance(NamespacedIdentifiableCollector.class, new NamespacedIdentifiableCollector());
    final String primaryGroupName = AuthorizationHelper.getPrimaryGroup(subject, storage).map(Group::getName).orElse("none");
    visitors.putInstance(ExecutionMetrics.QueryMetricsReporter.class, new ExecutionMetrics.QueryMetricsReporter(primaryGroupName));
    // Chain all Consumers
    Consumer<Visitable> consumerChain = QueryUtils.getNoOpEntryPoint();
    for (QueryVisitor visitor : visitors.values()) {
        consumerChain = consumerChain.andThen(visitor);
    }
    // Apply consumers to the query tree
    query.visit(consumerChain);
    query.authorize(subject, dataset, visitors);
    // After all authorization checks we can now use the actual subject to invoke the query and do not to bubble down the Userish in methods
    ExecutionMetrics.reportNamespacedIds(visitors.getInstance(NamespacedIdentifiableCollector.class).getIdentifiables(), primaryGroupName);
    ExecutionMetrics.reportQueryClassUsage(query.getClass(), primaryGroupName);
    final Namespace namespace = datasetRegistry.get(dataset.getId());
    final ExecutionManager executionManager = namespace.getExecutionManager();
    // If this is only a re-executing query, try to execute the underlying query instead.
    {
        final Optional<ManagedExecutionId> executionId = visitors.getInstance(QueryUtils.OnlyReusingChecker.class).getOnlyReused();
        final Optional<ManagedExecution<?>> execution = executionId.map(id -> tryReuse(query, id, datasetRegistry, config, executionManager, subject.getUser()));
        if (execution.isPresent()) {
            return execution.get();
        }
    }
    // Execute the query
    return executionManager.runQuery(datasetRegistry, query, subject.getUser(), dataset, config);
}
Also used : ManagedQuery(com.bakdata.conquery.models.query.ManagedQuery) ConqueryConfig(com.bakdata.conquery.models.config.ConqueryConfig) URL(java.net.URL) ExternalUpload(com.bakdata.conquery.apiv1.query.ExternalUpload) RequiredArgsConstructor(lombok.RequiredArgsConstructor) ManagedExecution(com.bakdata.conquery.models.execution.ManagedExecution) Map(java.util.Map) DatasetId(com.bakdata.conquery.models.identifiable.ids.specific.DatasetId) BadRequestException(javax.ws.rs.BadRequestException) UriBuilder(javax.ws.rs.core.UriBuilder) QueryDescription(com.bakdata.conquery.apiv1.query.QueryDescription) CQAnd(com.bakdata.conquery.apiv1.query.concept.specific.CQAnd) ExecutionMetrics(com.bakdata.conquery.metrics.ExecutionMetrics) Collection(java.util.Collection) SecondaryIdQuery(com.bakdata.conquery.apiv1.query.SecondaryIdQuery) Visitable(com.bakdata.conquery.models.query.Visitable) Set(java.util.Set) User(com.bakdata.conquery.models.auth.entities.User) Collectors(java.util.stream.Collectors) Dataset(com.bakdata.conquery.models.datasets.Dataset) ConceptQuery(com.bakdata.conquery.apiv1.query.ConceptQuery) ExternalUploadResult(com.bakdata.conquery.apiv1.query.ExternalUploadResult) SecondaryIdDescription(com.bakdata.conquery.models.datasets.SecondaryIdDescription) List(java.util.List) Slf4j(lombok.extern.slf4j.Slf4j) Stream(java.util.stream.Stream) Response(javax.ws.rs.core.Response) CancelQuery(com.bakdata.conquery.models.messages.namespaces.specific.CancelQuery) Ability(com.bakdata.conquery.models.auth.permissions.Ability) QueryUtils(com.bakdata.conquery.util.QueryUtils) Optional(java.util.Optional) MetaStorage(com.bakdata.conquery.io.storage.MetaStorage) Namespace(com.bakdata.conquery.models.worker.Namespace) ExecutionManager(com.bakdata.conquery.models.query.ExecutionManager) ResultRendererProvider(com.bakdata.conquery.io.result.ResultRender.ResultRendererProvider) Getter(lombok.Getter) ExecutionState(com.bakdata.conquery.models.execution.ExecutionState) Subject(com.bakdata.conquery.models.auth.entities.Subject) ClassToInstanceMap(com.google.common.collect.ClassToInstanceMap) HttpServletRequest(javax.servlet.http.HttpServletRequest) ManagedExecutionId(com.bakdata.conquery.models.identifiable.ids.specific.ManagedExecutionId) QueryVisitor(com.bakdata.conquery.models.query.visitor.QueryVisitor) NamespacedIdentifiableCollector(com.bakdata.conquery.util.QueryUtils.NamespacedIdentifiableCollector) Group(com.bakdata.conquery.models.auth.entities.Group) CQExternal(com.bakdata.conquery.apiv1.query.concept.specific.external.CQExternal) AuthorizationHelper(com.bakdata.conquery.models.auth.AuthorizationHelper) Consumer(java.util.function.Consumer) MutableClassToInstanceMap(com.google.common.collect.MutableClassToInstanceMap) AuthorizationHelper.buildDatasetAbilityMap(com.bakdata.conquery.models.auth.AuthorizationHelper.buildDatasetAbilityMap) DatasetRegistry(com.bakdata.conquery.models.worker.DatasetRegistry) CQElement(com.bakdata.conquery.apiv1.query.CQElement) Query(com.bakdata.conquery.apiv1.query.Query) QueryVisitor(com.bakdata.conquery.models.query.visitor.QueryVisitor) ExecutionManager(com.bakdata.conquery.models.query.ExecutionManager) NamespacedIdentifiableCollector(com.bakdata.conquery.util.QueryUtils.NamespacedIdentifiableCollector) Optional(java.util.Optional) Visitable(com.bakdata.conquery.models.query.Visitable) ExecutionMetrics(com.bakdata.conquery.metrics.ExecutionMetrics) Namespace(com.bakdata.conquery.models.worker.Namespace) QueryUtils(com.bakdata.conquery.util.QueryUtils)

Example 3 with ManagedExecution

use of com.bakdata.conquery.models.execution.ManagedExecution in project conquery by bakdata.

the class ResultExcelProcessor method getExcelResult.

public <E extends ManagedExecution<?> & SingleTableResult> Response getExcelResult(Subject subject, E exec, DatasetId datasetId, boolean pretty) {
    ConqueryMDC.setLocation(subject.getName());
    final Namespace namespace = datasetRegistry.get(datasetId);
    Dataset dataset = namespace.getDataset();
    subject.authorize(dataset, Ability.READ);
    subject.authorize(dataset, Ability.DOWNLOAD);
    subject.authorize(exec, Ability.READ);
    IdPrinter idPrinter = config.getFrontend().getQueryUpload().getIdPrinter(subject, exec, namespace);
    final Locale locale = I18n.LOCALE.get();
    PrintSettings settings = new PrintSettings(pretty, locale, datasetRegistry, config, idPrinter::createId);
    ExcelRenderer excelRenderer = new ExcelRenderer(config.getExcel(), settings);
    StreamingOutput out = output -> excelRenderer.renderToStream(config.getFrontend().getQueryUpload().getIdResultInfos(), (ManagedExecution<?> & SingleTableResult) exec, output);
    return makeResponseWithFileName(out, exec.getLabelWithoutAutoLabelSuffix(), "xlsx", MEDIA_TYPE, ResultUtil.ContentDispositionOption.ATTACHMENT);
}
Also used : IdPrinter(com.bakdata.conquery.models.identifiable.mapping.IdPrinter) Locale(java.util.Locale) ResultUtil(com.bakdata.conquery.io.result.ResultUtil) ConqueryMDC(com.bakdata.conquery.util.io.ConqueryMDC) ConqueryConfig(com.bakdata.conquery.models.config.ConqueryConfig) ResultUtil.makeResponseWithFileName(com.bakdata.conquery.io.result.ResultUtil.makeResponseWithFileName) Subject(com.bakdata.conquery.models.auth.entities.Subject) RequiredArgsConstructor(lombok.RequiredArgsConstructor) StreamingOutput(javax.ws.rs.core.StreamingOutput) SingleTableResult(com.bakdata.conquery.models.query.SingleTableResult) ArrayList(java.util.ArrayList) Dataset(com.bakdata.conquery.models.datasets.Dataset) PrintSettings(com.bakdata.conquery.models.query.PrintSettings) List(java.util.List) MediaType(javax.ws.rs.core.MediaType) Response(javax.ws.rs.core.Response) ManagedExecution(com.bakdata.conquery.models.execution.ManagedExecution) Ability(com.bakdata.conquery.models.auth.permissions.Ability) Locale(java.util.Locale) DatasetId(com.bakdata.conquery.models.identifiable.ids.specific.DatasetId) I18n(com.bakdata.conquery.models.i18n.I18n) DatasetRegistry(com.bakdata.conquery.models.worker.DatasetRegistry) IdPrinter(com.bakdata.conquery.models.identifiable.mapping.IdPrinter) ResultInfo(com.bakdata.conquery.models.query.resultinfo.ResultInfo) Namespace(com.bakdata.conquery.models.worker.Namespace) Dataset(com.bakdata.conquery.models.datasets.Dataset) PrintSettings(com.bakdata.conquery.models.query.PrintSettings) StreamingOutput(javax.ws.rs.core.StreamingOutput) SingleTableResult(com.bakdata.conquery.models.query.SingleTableResult) ManagedExecution(com.bakdata.conquery.models.execution.ManagedExecution) Namespace(com.bakdata.conquery.models.worker.Namespace)

Example 4 with ManagedExecution

use of com.bakdata.conquery.models.execution.ManagedExecution in project conquery by bakdata.

the class TestConquery method isBusy.

private boolean isBusy() {
    boolean busy;
    busy = standaloneCommand.getManager().getJobManager().isSlowWorkerBusy();
    busy |= standaloneCommand.getManager().getStorage().getAllExecutions().stream().map(ManagedExecution::getState).anyMatch(ExecutionState.RUNNING::equals);
    for (Namespace namespace : standaloneCommand.getManager().getDatasetRegistry().getDatasets()) {
        busy |= namespace.getJobManager().isSlowWorkerBusy();
    }
    for (ShardNode slave : standaloneCommand.getShardNodes()) {
        busy |= slave.isBusy();
    }
    return busy;
}
Also used : ExecutionState(com.bakdata.conquery.models.execution.ExecutionState) ShardNode(com.bakdata.conquery.commands.ShardNode) ManagedExecution(com.bakdata.conquery.models.execution.ManagedExecution) Namespace(com.bakdata.conquery.models.worker.Namespace)

Example 5 with ManagedExecution

use of com.bakdata.conquery.models.execution.ManagedExecution in project conquery by bakdata.

the class ResultArrowProcessor method getArrowResult.

public static <E extends ManagedExecution<?> & SingleTableResult> Response getArrowResult(Function<OutputStream, Function<VectorSchemaRoot, ArrowWriter>> writerProducer, Subject subject, E exec, Dataset dataset, DatasetRegistry datasetRegistry, boolean pretty, String fileExtension, MediaType mediaType, ConqueryConfig config) {
    final Namespace namespace = datasetRegistry.get(dataset.getId());
    ConqueryMDC.setLocation(subject.getName());
    log.info("Downloading results for {} on dataset {}", exec, dataset);
    subject.authorize(dataset, Ability.READ);
    subject.authorize(dataset, Ability.DOWNLOAD);
    subject.authorize(exec, Ability.READ);
    // Check if subject is permitted to download on all datasets that were referenced by the query
    authorizeDownloadDatasets(subject, exec);
    if (!(exec instanceof ManagedQuery || (exec instanceof ManagedForm && ((ManagedForm) exec).getSubQueries().size() == 1))) {
        return Response.status(HttpStatus.SC_UNPROCESSABLE_ENTITY, "Execution result is not a single Table").build();
    }
    // Get the locale extracted by the LocaleFilter
    IdPrinter idPrinter = config.getFrontend().getQueryUpload().getIdPrinter(subject, exec, namespace);
    final Locale locale = I18n.LOCALE.get();
    PrintSettings settings = new PrintSettings(pretty, locale, datasetRegistry, config, idPrinter::createId);
    // Collect ResultInfos for id columns and result columns
    final List<ResultInfo> resultInfosId = config.getFrontend().getQueryUpload().getIdResultInfos();
    final List<ResultInfo> resultInfosExec = exec.getResultInfos();
    StreamingOutput out = output -> renderToStream(writerProducer.apply(output), settings, config.getArrow().getBatchSize(), resultInfosId, resultInfosExec, exec.streamResults());
    return makeResponseWithFileName(out, exec.getLabelWithoutAutoLabelSuffix(), fileExtension, mediaType, ResultUtil.ContentDispositionOption.ATTACHMENT);
}
Also used : IdPrinter(com.bakdata.conquery.models.identifiable.mapping.IdPrinter) Locale(java.util.Locale) ManagedQuery(com.bakdata.conquery.models.query.ManagedQuery) ConqueryConfig(com.bakdata.conquery.models.config.ConqueryConfig) ManagedForm(com.bakdata.conquery.models.forms.managed.ManagedForm) Subject(com.bakdata.conquery.models.auth.entities.Subject) AuthorizationHelper.authorizeDownloadDatasets(com.bakdata.conquery.models.auth.AuthorizationHelper.authorizeDownloadDatasets) HttpStatus(org.apache.http.HttpStatus) Function(java.util.function.Function) ArrayList(java.util.ArrayList) PrintSettings(com.bakdata.conquery.models.query.PrintSettings) UtilityClass(lombok.experimental.UtilityClass) MediaType(javax.ws.rs.core.MediaType) ManagedExecution(com.bakdata.conquery.models.execution.ManagedExecution) ArrowWriter(org.apache.arrow.vector.ipc.ArrowWriter) Locale(java.util.Locale) ArrowRenderer.renderToStream(com.bakdata.conquery.io.result.arrow.ArrowRenderer.renderToStream) I18n(com.bakdata.conquery.models.i18n.I18n) IdPrinter(com.bakdata.conquery.models.identifiable.mapping.IdPrinter) ResultInfo(com.bakdata.conquery.models.query.resultinfo.ResultInfo) OutputStream(java.io.OutputStream) ResultUtil(com.bakdata.conquery.io.result.ResultUtil) ConqueryMDC(com.bakdata.conquery.util.io.ConqueryMDC) ResultUtil.makeResponseWithFileName(com.bakdata.conquery.io.result.ResultUtil.makeResponseWithFileName) VectorSchemaRoot(org.apache.arrow.vector.VectorSchemaRoot) StreamingOutput(javax.ws.rs.core.StreamingOutput) SingleTableResult(com.bakdata.conquery.models.query.SingleTableResult) Dataset(com.bakdata.conquery.models.datasets.Dataset) List(java.util.List) Slf4j(lombok.extern.slf4j.Slf4j) Response(javax.ws.rs.core.Response) Ability(com.bakdata.conquery.models.auth.permissions.Ability) DatasetRegistry(com.bakdata.conquery.models.worker.DatasetRegistry) Namespace(com.bakdata.conquery.models.worker.Namespace) ManagedForm(com.bakdata.conquery.models.forms.managed.ManagedForm) PrintSettings(com.bakdata.conquery.models.query.PrintSettings) StreamingOutput(javax.ws.rs.core.StreamingOutput) ManagedQuery(com.bakdata.conquery.models.query.ManagedQuery) ResultInfo(com.bakdata.conquery.models.query.resultinfo.ResultInfo) Namespace(com.bakdata.conquery.models.worker.Namespace)

Aggregations

ManagedExecution (com.bakdata.conquery.models.execution.ManagedExecution)6 Namespace (com.bakdata.conquery.models.worker.Namespace)5 Subject (com.bakdata.conquery.models.auth.entities.Subject)4 Ability (com.bakdata.conquery.models.auth.permissions.Ability)4 ConqueryConfig (com.bakdata.conquery.models.config.ConqueryConfig)4 Dataset (com.bakdata.conquery.models.datasets.Dataset)4 DatasetRegistry (com.bakdata.conquery.models.worker.DatasetRegistry)4 List (java.util.List)4 Response (javax.ws.rs.core.Response)4 Slf4j (lombok.extern.slf4j.Slf4j)4 ResultUtil (com.bakdata.conquery.io.result.ResultUtil)3 ResultUtil.makeResponseWithFileName (com.bakdata.conquery.io.result.ResultUtil.makeResponseWithFileName)3 ManagedQuery (com.bakdata.conquery.models.query.ManagedQuery)3 RequiredArgsConstructor (lombok.RequiredArgsConstructor)3 MetaStorage (com.bakdata.conquery.io.storage.MetaStorage)2 AuthorizationHelper.authorizeDownloadDatasets (com.bakdata.conquery.models.auth.AuthorizationHelper.authorizeDownloadDatasets)2 ExecutionState (com.bakdata.conquery.models.execution.ExecutionState)2 I18n (com.bakdata.conquery.models.i18n.I18n)2 DatasetId (com.bakdata.conquery.models.identifiable.ids.specific.DatasetId)2 IdPrinter (com.bakdata.conquery.models.identifiable.mapping.IdPrinter)2