Examples with DataSource com.thinkbiganalytics.metadata.api.catalog.DataSource used on opensource projects

Example 6 with DataSource

use of com.thinkbiganalytics.metadata.api.catalog.DataSource in project kylo by Teradata.

the class AccessControlConfigurator method ensureCatalogAccessControl.

/**
 * Ensures that the entity-level access control is setup up for the entities introduced by the connector architecture.
 */
public void ensureCatalogAccessControl() {
    List<Connector> connectors = connectorProvider.findAll(true);
    List<SecurityRole> conntorRoles = this.roleProvider.getEntityRoles(SecurityRole.CONNECTOR);
    Optional<AllowedActions> connectorActions = this.actionsProvider.getAvailableActions(AllowedActions.CONNECTOR);
    connectors.stream().forEach(conn -> {
        Principal owner = conn.getOwner() != null ? conn.getOwner() : JcrMetadataAccess.getActiveUser();
        connectorActions.ifPresent(actions -> ((JcrConnector) conn).enableAccessControl((JcrAllowedActions) actions, owner, conntorRoles));
    });
    List<DataSource> dataSources = dataSourceProvider.findAll();
    List<SecurityRole> dataSourceRoles = this.roleProvider.getEntityRoles(SecurityRole.DATASOURCE);
    Optional<AllowedActions> dataSourceActions = this.actionsProvider.getAvailableActions(AllowedActions.DATASOURCE);
    dataSources.stream().map(JcrDataSource.class::cast).forEach(dataSource -> {
        Principal owner = dataSource.getOwner() != null ? dataSource.getOwner() : JcrMetadataAccess.getActiveUser();
        dataSourceActions.ifPresent(actions -> dataSource.enableAccessControl((JcrAllowedActions) actions, owner, dataSourceRoles));
    });
}

Example 7 with DataSource

use of com.thinkbiganalytics.metadata.api.catalog.DataSource in project kylo by Teradata.

the class DefaultFeedManagerFeedService method assignFeedDatasources.

/**
 * Assign the feed sources/destinations
 *
 * @param feed       the feed rest model
 * @param domainFeed the domain feed
 */
private void assignFeedDatasources(FeedMetadata feed, Feed domainFeed) {
    final Feed.ID domainFeedId = domainFeed.getId();
    Set<com.thinkbiganalytics.metadata.api.catalog.DataSet.ID> sourceDataSets = new HashSet<>();
    Set<com.thinkbiganalytics.metadata.api.datasource.Datasource.ID> sourceDatasources = new HashSet<>();
    Set<com.thinkbiganalytics.metadata.api.datasource.Datasource.ID> destinationDatasources = new HashSet<>();
    String uniqueName = FeedNameUtil.fullName(feed.getCategory().getSystemName(), feed.getSystemFeedName());
    RegisteredTemplate template = feed.getRegisteredTemplate();
    if (template == null) {
        // fetch it for checks
        template = templateRestProvider.getRegisteredTemplate(feed.getTemplateId());
    }
    // Collect the IDs of the legacy datasources the feed had referenced
    Set<Datasource.ID> previousSourceIds = domainFeed.getSources().stream().filter(fs -> fs.getDatasource().isPresent()).map(fs -> fs.getDatasource().get().getId()).collect(Collectors.toSet());
    Set<Datasource.ID> previousDestIds = domainFeed.getDestinations().stream().filter(// Currently will always be true as there are no destination data sets yet.
    fs -> fs.getDatasource().isPresent()).map(fs -> fs.getDatasource().get().getId()).collect(Collectors.toSet());
    boolean isSampleDataSet = isTreatSourceDataSetsAsSample(feed, template);
    // find Definition registration
    derivedDatasourceFactory.populateDatasources(feed, template, sourceDatasources, destinationDatasources);
    // Replace the older legacy datasource references with the new ones.
    previousSourceIds.stream().filter(id -> !sourceDatasources.contains(id)).forEach(id -> feedProvider.removeFeedSource(domainFeedId, id));
    sourceDatasources.stream().forEach(sourceId -> feedProvider.ensureFeedSource(domainFeedId, sourceId));
    previousDestIds.stream().filter(id -> !destinationDatasources.contains(id)).forEach(id -> feedProvider.removeFeedDestination(domainFeedId, id));
    destinationDatasources.stream().forEach(sourceId -> feedProvider.ensureFeedDestination(domainFeedId, sourceId));
    // Update data sets
    if (feed.getSourceDataSets() != null) {
        // Collect the IDs of source data sets the feed had referenced
        Set<com.thinkbiganalytics.metadata.api.catalog.DataSet.ID> currentDataSetIds = domainFeed.getSources().stream().map(FeedSource::getDataSet).filter(Optional::isPresent).map(Optional::get).map(com.thinkbiganalytics.metadata.api.catalog.DataSet::getId).collect(Collectors.toSet());
        Set<com.thinkbiganalytics.metadata.api.catalog.DataSet.ID> newDataSetIds = new HashSet<>();
        feed.getSourceDataSets().forEach(dataSet -> {
            com.thinkbiganalytics.metadata.api.catalog.DataSet addedDataSet;
            if (dataSet.getId() == null) {
                DataSource.ID dataSourceId = dataSourceProvider.resolveId(dataSet.getDataSource().getId());
                dataSourceProvider.find(dataSourceId).orElseThrow(() -> new DataSourceNotFoundException(dataSourceId));
                addedDataSet = catalogModelTransform.buildDataSet(dataSet, dataSetProvider.build(dataSourceId));
            } else {
                com.thinkbiganalytics.metadata.api.catalog.DataSet.ID dataSetId = dataSetProvider.resolveId(dataSet.getId());
                addedDataSet = dataSetProvider.find(dataSetId).orElseThrow(() -> new DataSetNotFoundException(dataSetId));
            }
            newDataSetIds.add(addedDataSet.getId());
            catalogModelTransform.updateDataSet(dataSet, addedDataSet);
            feedProvider.ensureFeedSource(domainFeedId, addedDataSet.getId(), isSampleDataSet);
        });
        // Remove any data set sources no longer referenced in the updated feed.
        currentDataSetIds.stream().filter(id -> !newDataSetIds.contains(id)).forEach(id -> feedProvider.removeFeedSource(domainFeedId, id));
    }
}

Example 8 with DataSource

use of com.thinkbiganalytics.metadata.api.catalog.DataSource in project kylo by Teradata.

the class DefaultFeedManagerFeedService method saveDeployedFeed.

private void saveDeployedFeed(final FeedMetadata feed, com.thinkbiganalytics.metadata.api.versioning.EntityVersion<Feed.ID, Feed> version) {
    Stopwatch stopwatch = Stopwatch.createStarted();
    List<? extends HadoopSecurityGroup> previousSavedSecurityGroups = null;
    // Store the old security groups before saving because we need to compare afterward
    Feed previousStateBeforeSaving = feedProvider.findById(feedProvider.resolveId(feed.getId()));
    Map<String, String> userProperties = previousStateBeforeSaving.getUserProperties();
    previousSavedSecurityGroups = previousStateBeforeSaving.getSecurityGroups();
    // if this is the first time saving this feed create a new one
    Feed domainFeed = version.getEntity().get();
    Feed.ID domainId = domainFeed.getId();
    if (domainFeed.getState() == null) {
        domainFeed.setState(Feed.State.ENABLED);
    } else if (Feed.State.ENABLED.name().contentEquals(feed.getState())) {
        domainFeed.setState(Feed.State.ENABLED);
    } else {
        domainFeed.setState(Feed.State.DISABLED);
    }
    stopwatch.stop();
    log.debug("Time to transform the feed to a domain object for saving: {} ms", stopwatch.elapsed(TimeUnit.MILLISECONDS));
    stopwatch.reset();
    final String feedName = FeedNameUtil.fullName(domainFeed.getCategory().getSystemName(), domainFeed.getName());
    stopwatch.start();
    boolean isStream = feed.getRegisteredTemplate() != null ? feed.getRegisteredTemplate().isStream() : false;
    Long timeBetweenBatchJobs = feed.getRegisteredTemplate() != null ? feed.getRegisteredTemplate().getTimeBetweenStartingBatchJobs() : 0L;
    // sync the feed information to ops manager
    metadataAccess.commit(() -> opsManagerFeedProvider.save(opsManagerFeedProvider.resolveId(domainId.toString()), feedName, isStream, timeBetweenBatchJobs));
    stopwatch.stop();
    log.debug("Time to sync feed data with Operations Manager: {} ms", stopwatch.elapsed(TimeUnit.MILLISECONDS));
    stopwatch.reset();
    // Update hadoop security group polices if the groups changed
    if (!ListUtils.isEqualList(previousSavedSecurityGroups, domainFeed.getSecurityGroups())) {
        stopwatch.start();
        List<? extends HadoopSecurityGroup> securityGroups = domainFeed.getSecurityGroups();
        List<String> groupsAsCommaList = securityGroups.stream().map(group -> group.getName()).collect(Collectors.toList());
        hadoopAuthorizationService.updateSecurityGroupsForAllPolicies(feed.getSystemCategoryName(), feed.getSystemFeedName(), groupsAsCommaList, domainFeed.getProperties());
        stopwatch.stop();
        log.debug("Time to update hadoop security groups: {} ms", stopwatch.elapsed(TimeUnit.MILLISECONDS));
        stopwatch.reset();
    }
    // Update Hive metastore
    if (hiveTargetSyncColumnDescriptions) {
        stopwatch.start();
        final boolean hasHiveDestination = domainFeed.getDestinations().stream().map(FeedDestination::getDatasource).filter(DerivedDatasource.class::isInstance).map(DerivedDatasource.class::cast).anyMatch(datasource -> "HiveDatasource".equals(datasource.getDatasourceType()));
        if (hasHiveDestination) {
            try {
                feedHiveTableService.updateColumnDescriptions(feed);
            } catch (final DataAccessException e) {
                log.warn("Failed to update column descriptions for feed: {}", feed.getCategoryAndFeedDisplayName(), e);
            }
        }
        stopwatch.stop();
        log.debug("Time to update hive metastore: {} ms", stopwatch.elapsed(TimeUnit.MILLISECONDS));
        stopwatch.reset();
    }
    // Update Kylo metastore
    stopwatch.start();
    feedProvider.setDeployed(domainId, version.getId());
    stopwatch.stop();
    log.debug("Time to call feedProvider.update: {} ms", stopwatch.elapsed(TimeUnit.MILLISECONDS));
    stopwatch.reset();
}