Examples with Datasource com.thinkbiganalytics.metadata.api.datasource.Datasource used on opensource projects

Example 1 with Datasource

use of com.thinkbiganalytics.metadata.api.datasource.Datasource in project kylo by Teradata.

the class FeedManagerFeedTest method testFeedDatasource.

@Test
public void testFeedDatasource() {
    String categorySystemName = "my_category";
    String feedName = "my_feed";
    String templateName = "my_template";
    String description = " my feed description";
    setupFeedAndTemplate(categorySystemName, feedName, templateName);
    // boolean isDefineTable = true;
    // boolean isGetFile = false;
    metadata.commit(() -> {
        Feed feed = feedTestUtil.findFeed(categorySystemName, feedName);
        Set<Datasource.ID> sources = new HashSet<Datasource.ID>();
        Set<com.thinkbiganalytics.metadata.api.datasource.Datasource.ID> destinations = new HashSet<>();
        // Add Table Dependencies
        String uniqueName = FeedNameUtil.fullName(categorySystemName, feedName);
        DerivedDatasource srcDatasource = datasourceProvider.ensureDatasource(uniqueName, feed.getDescription(), DerivedDatasource.class);
        sources.add(srcDatasource.getId());
        DerivedDatasource destDatasource = datasourceProvider.ensureDatasource("destination", feed.getDescription(), DerivedDatasource.class);
        destinations.add(destDatasource.getId());
        sources.stream().forEach(sourceId -> feedProvider.ensureFeedSource(feed.getId(), sourceId));
        destinations.stream().forEach(destinationId -> feedProvider.ensureFeedDestination(feed.getId(), destinationId));
    }, MetadataAccess.SERVICE);
    // ensure the sources and dest got created
    metadata.read(() -> {
        Feed feed = feedTestUtil.findFeed(categorySystemName, feedName);
        Assert.assertNotNull(feed.getSources());
        Assert.assertTrue(feed.getSources().size() == 1, "Feed Sources should be 1");
        Assert.assertNotNull(feed.getDestinations());
        Assert.assertTrue(feed.getDestinations().size() == 1, "Feed Destinations should be 1");
        List<? extends FeedDestination> feedDestinations = feed.getDestinations();
        if (feedDestinations != null) {
            FeedDestination feedDestination = feedDestinations.get(0);
            Datasource ds = feedDestination.getDatasource();
            Assert.assertTrue(ds instanceof DerivedDatasource, "Datasource was not expected DerivedDatasource");
        }
    }, MetadataAccess.SERVICE);
}

Example 2 with Datasource

use of com.thinkbiganalytics.metadata.api.datasource.Datasource in project kylo by Teradata.

the class InMemoryFeedProvider method ensureFeed.

@Override
public Feed ensureFeed(String categorySystemName, String name, String descr, ID srcId, ID destId) {
    Datasource sds = this.datasetProvider.getDatasource(srcId);
    Datasource dds = this.datasetProvider.getDatasource(destId);
    if (sds == null) {
        throw new FeedCreateException("A dataset with the given ID does not exists: " + srcId);
    }
    if (dds == null) {
        throw new FeedCreateException("A dataset with the given ID does not exists: " + destId);
    }
    BaseFeed feed = (BaseFeed) ensureFeed(categorySystemName, name, descr);
    ensureFeedSource(feed, sds, null);
    ensureFeedDestination(feed, dds);
    return feed;
}

Example 3 with Datasource

use of com.thinkbiganalytics.metadata.api.datasource.Datasource in project kylo by Teradata.

the class InMemoryFeedProvider method ensureFeedSource.

@Override
public FeedSource ensureFeedSource(Feed.ID feedId, Datasource.ID dsId, ServiceLevelAgreement.ID slaId) {
    BaseFeed feed = (BaseFeed) this.feeds.get(feedId);
    Datasource ds = this.datasetProvider.getDatasource(dsId);
    if (feed == null) {
        throw new FeedCreateException("A feed with the given ID does not exists: " + feedId);
    }
    if (ds == null) {
        throw new FeedCreateException("A dataset with the given ID does not exists: " + dsId);
    }
    return ensureFeedSource(feed, ds, slaId);
}

Example 4 with Datasource

use of com.thinkbiganalytics.metadata.api.datasource.Datasource in project kylo by Teradata.

the class DefaultFeedManagerFeedService method saveFeed.

private void saveFeed(final FeedMetadata feed) {
    metadataAccess.commit(() -> {
        Stopwatch stopwatch = Stopwatch.createStarted();
        List<? extends HadoopSecurityGroup> previousSavedSecurityGroups = null;
        // Store the old security groups before saving because we need to compare afterward
        if (!feed.isNew()) {
            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 = feedModelTransform.feedToDomain(feed);
        if (domainFeed.getState() == null) {
            domainFeed.setState(Feed.State.ENABLED);
        }
        stopwatch.stop();
        log.debug("Time to transform the feed to a domain object for saving: {} ms", stopwatch.elapsed(TimeUnit.MILLISECONDS));
        stopwatch.reset();
        // initially save the feed
        if (feed.isNew()) {
            stopwatch.start();
            domainFeed = feedProvider.update(domainFeed);
            stopwatch.stop();
            log.debug("Time to save the New feed: {} ms", stopwatch.elapsed(TimeUnit.MILLISECONDS));
            stopwatch.reset();
        }
        final String domainId = domainFeed.getId().toString();
        final String feedName = FeedNameUtil.fullName(domainFeed.getCategory().getSystemName(), domainFeed.getName());
        // Build preconditions
        stopwatch.start();
        assignFeedDependencies(feed, domainFeed);
        stopwatch.stop();
        log.debug("Time to assignFeedDependencies: {} ms", stopwatch.elapsed(TimeUnit.MILLISECONDS));
        stopwatch.reset();
        // Assign the datasources
        stopwatch.start();
        assignFeedDatasources(feed, domainFeed);
        stopwatch.stop();
        log.debug("Time to assignFeedDatasources: {} ms", stopwatch.elapsed(TimeUnit.MILLISECONDS));
        stopwatch.reset();
        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), 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 (!feed.isNew() && !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
        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();
        domainFeed = feedProvider.update(domainFeed);
        stopwatch.stop();
        log.debug("Time to call feedProvider.update: {} ms", stopwatch.elapsed(TimeUnit.MILLISECONDS));
        stopwatch.reset();
    }, (e) -> {
        if (feed.isNew() && StringUtils.isNotBlank(feed.getId())) {
            // Rollback ops Manager insert if it is newly created
            metadataAccess.commit(() -> {
                opsManagerFeedProvider.delete(opsManagerFeedProvider.resolveId(feed.getId()));
            });
        }
    });
}

Example 5 with Datasource

use of com.thinkbiganalytics.metadata.api.datasource.Datasource in project kylo by Teradata.

the class DerivedDatasourceFactory method ensureDataTransformationSourceDatasources.

/**
 * Builds the list of data sources for the specified data transformation feed.
 *
 * @param feed the feed
 * @return the list of data sources
 * @throws NullPointerException if the feed has no data transformation
 */
@Nonnull
private Set<Datasource.ID> ensureDataTransformationSourceDatasources(@Nonnull final FeedMetadata feed) {
    final Set<Datasource.ID> datasources = new HashSet<>();
    // Extract nodes in chart view model
    @SuppressWarnings("unchecked") final Stream<Map<String, Object>> nodes = Optional.ofNullable(feed.getDataTransformation().getChartViewModel()).map(model -> (List<Map<String, Object>>) model.get("nodes")).map(Collection::stream).orElse(Stream.empty());
    // Create a data source for each node
    final DatasourceDefinition hiveDefinition = datasourceDefinitionProvider.findByProcessorType(DATA_TRANSFORMATION_HIVE_DEFINITION);
    final DatasourceDefinition jdbcDefinition = datasourceDefinitionProvider.findByProcessorType(DATA_TRANSFORMATION_JDBC_DEFINITION);
    nodes.forEach(node -> {
        // Extract properties from node
        final DatasourceDefinition datasourceDefinition;
        final Map<String, String> properties = new HashMap<>();
        if (node.get("datasourceId") == null || node.get("datasourceId").equals("HIVE")) {
            final String name = (String) node.get("name");
            datasourceDefinition = hiveDefinition;
            properties.put(HIVE_SCHEMA_KEY, StringUtils.trim(StringUtils.substringBefore(name, ".")));
            properties.put(HIVE_TABLE_KEY, StringUtils.trim(StringUtils.substringAfterLast(name, ".")));
        } else {
            final Datasource datasource = datasourceProvider.getDatasource(datasourceProvider.resolve((String) node.get("datasourceId")));
            datasourceDefinition = jdbcDefinition;
            properties.put(JDBC_CONNECTION_KEY, datasource.getName());
            properties.put(JDBC_TABLE_KEY, (String) node.get("name"));
            properties.putAll(parseDataTransformControllerServiceProperties(datasourceDefinition, datasource.getName()));
        }
        if (datasourceDefinition != null) {
            // Create the derived data source
            final String identityString = propertyExpressionResolver.resolveVariables(datasourceDefinition.getIdentityString(), properties);
            final String title = datasourceDefinition.getTitle() != null ? propertyExpressionResolver.resolveVariables(datasourceDefinition.getTitle(), properties) : identityString;
            final String desc = propertyExpressionResolver.resolveVariables(datasourceDefinition.getDescription(), properties);
            final DerivedDatasource datasource = datasourceProvider.ensureDerivedDatasource(datasourceDefinition.getDatasourceType(), identityString, title, desc, new HashMap<>(properties));
            datasources.add(datasource.getId());
        }
    });
    // Build the data sources from the data source ids
    final List<String> datasourceIds = Optional.ofNullable(feed.getDataTransformation()).map(FeedDataTransformation::getDatasourceIds).orElse(Collections.emptyList());
    datasourceIds.stream().map(datasourceProvider::resolve).forEach(datasources::add);
    return datasources;
}