use of io.druid.query.QueryRunner in project druid by druid-io.
the class SegmentMetadataQueryRunnerFactory method createRunner.
@Override
public QueryRunner<SegmentAnalysis> createRunner(final Segment segment) {
return new QueryRunner<SegmentAnalysis>() {
@Override
public Sequence<SegmentAnalysis> run(Query<SegmentAnalysis> inQ, Map<String, Object> responseContext) {
SegmentMetadataQuery query = (SegmentMetadataQuery) inQ;
final SegmentAnalyzer analyzer = new SegmentAnalyzer(query.getAnalysisTypes());
final Map<String, ColumnAnalysis> analyzedColumns = analyzer.analyze(segment);
final long numRows = analyzer.numRows(segment);
long totalSize = 0;
if (analyzer.analyzingSize()) {
// Initialize with the size of the whitespace, 1 byte per
totalSize = analyzedColumns.size() * numRows;
}
Map<String, ColumnAnalysis> columns = Maps.newTreeMap();
ColumnIncluderator includerator = query.getToInclude();
for (Map.Entry<String, ColumnAnalysis> entry : analyzedColumns.entrySet()) {
final String columnName = entry.getKey();
final ColumnAnalysis column = entry.getValue();
if (!column.isError()) {
totalSize += column.getSize();
}
if (includerator.include(columnName)) {
columns.put(columnName, column);
}
}
List<Interval> retIntervals = query.analyzingInterval() ? Arrays.asList(segment.getDataInterval()) : null;
final Map<String, AggregatorFactory> aggregators;
Metadata metadata = null;
if (query.hasAggregators()) {
metadata = segment.asStorageAdapter().getMetadata();
if (metadata != null && metadata.getAggregators() != null) {
aggregators = Maps.newHashMap();
for (AggregatorFactory aggregator : metadata.getAggregators()) {
aggregators.put(aggregator.getName(), aggregator);
}
} else {
aggregators = null;
}
} else {
aggregators = null;
}
final TimestampSpec timestampSpec;
if (query.hasTimestampSpec()) {
if (metadata == null) {
metadata = segment.asStorageAdapter().getMetadata();
}
timestampSpec = metadata != null ? metadata.getTimestampSpec() : null;
} else {
timestampSpec = null;
}
final Granularity queryGranularity;
if (query.hasQueryGranularity()) {
if (metadata == null) {
metadata = segment.asStorageAdapter().getMetadata();
}
queryGranularity = metadata != null ? metadata.getQueryGranularity() : null;
} else {
queryGranularity = null;
}
Boolean rollup = null;
if (query.hasRollup()) {
if (metadata == null) {
metadata = segment.asStorageAdapter().getMetadata();
}
rollup = metadata != null ? metadata.isRollup() : null;
if (rollup == null) {
// in this case, this segment is built before no-rollup function is coded,
// thus it is built with rollup
rollup = Boolean.TRUE;
}
}
return Sequences.simple(Arrays.asList(new SegmentAnalysis(segment.getIdentifier(), retIntervals, columns, totalSize, numRows, aggregators, timestampSpec, queryGranularity, rollup)));
}
};
}
use of io.druid.query.QueryRunner in project druid by druid-io.
the class SegmentMetadataQueryRunnerFactory method mergeRunners.
@Override
public QueryRunner<SegmentAnalysis> mergeRunners(ExecutorService exec, Iterable<QueryRunner<SegmentAnalysis>> queryRunners) {
final ListeningExecutorService queryExecutor = MoreExecutors.listeningDecorator(exec);
return new ConcatQueryRunner<SegmentAnalysis>(Sequences.map(Sequences.simple(queryRunners), new Function<QueryRunner<SegmentAnalysis>, QueryRunner<SegmentAnalysis>>() {
@Override
public QueryRunner<SegmentAnalysis> apply(final QueryRunner<SegmentAnalysis> input) {
return new QueryRunner<SegmentAnalysis>() {
@Override
public Sequence<SegmentAnalysis> run(final Query<SegmentAnalysis> query, final Map<String, Object> responseContext) {
final int priority = BaseQuery.getContextPriority(query, 0);
ListenableFuture<Sequence<SegmentAnalysis>> future = queryExecutor.submit(new AbstractPrioritizedCallable<Sequence<SegmentAnalysis>>(priority) {
@Override
public Sequence<SegmentAnalysis> call() throws Exception {
return Sequences.simple(Sequences.toList(input.run(query, responseContext), new ArrayList<SegmentAnalysis>()));
}
});
try {
queryWatcher.registerQuery(query, future);
final Number timeout = query.getContextValue(QueryContextKeys.TIMEOUT, (Number) null);
return timeout == null ? future.get() : future.get(timeout.longValue(), TimeUnit.MILLISECONDS);
} catch (InterruptedException e) {
log.warn(e, "Query interrupted, cancelling pending results, query id [%s]", query.getId());
future.cancel(true);
throw new QueryInterruptedException(e);
} catch (CancellationException e) {
throw new QueryInterruptedException(e);
} catch (TimeoutException e) {
log.info("Query timeout, cancelling pending results for query id [%s]", query.getId());
future.cancel(true);
throw new QueryInterruptedException(e);
} catch (ExecutionException e) {
throw Throwables.propagate(e.getCause());
}
}
};
}
}));
}
use of io.druid.query.QueryRunner in project druid by druid-io.
the class CachingClusteredClientTest method testTopNCachingTimeZone.
@Test
@SuppressWarnings("unchecked")
public void testTopNCachingTimeZone() throws Exception {
final TopNQueryBuilder builder = new TopNQueryBuilder().dataSource(DATA_SOURCE).dimension(TOP_DIM).metric("imps").threshold(3).intervals(SEG_SPEC).filters(DIM_FILTER).granularity(PT1H_TZ_GRANULARITY).aggregators(AGGS).postAggregators(POST_AGGS).context(CONTEXT);
QueryRunner runner = new FinalizeResultsQueryRunner(client, new TopNQueryQueryToolChest(new TopNQueryConfig(), QueryRunnerTestHelper.NoopIntervalChunkingQueryRunnerDecorator()));
testQueryCaching(runner, builder.build(), new Interval("2011-11-04/2011-11-08"), makeTopNResultsWithoutRename(new DateTime("2011-11-04", TIMEZONE), "a", 50, 4994, "b", 50, 4993, "c", 50, 4992, new DateTime("2011-11-05", TIMEZONE), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, new DateTime("2011-11-06", TIMEZONE), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, new DateTime("2011-11-07", TIMEZONE), "a", 50, 4988, "b", 50, 4987, "c", 50, 4986));
HashMap<String, List> context = new HashMap<String, List>();
TestHelper.assertExpectedResults(makeRenamedTopNResults(new DateTime("2011-11-04", TIMEZONE), "a", 50, 4994, "b", 50, 4993, "c", 50, 4992, new DateTime("2011-11-05", TIMEZONE), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, new DateTime("2011-11-06", TIMEZONE), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, new DateTime("2011-11-07", TIMEZONE), "a", 50, 4988, "b", 50, 4987, "c", 50, 4986), runner.run(builder.intervals("2011-11-04/2011-11-08").metric("imps").aggregators(RENAMED_AGGS).postAggregators(DIFF_ORDER_POST_AGGS).build(), context));
}
use of io.druid.query.QueryRunner in project druid by druid-io.
the class CachingClusteredClientTest method testTopNCachingEmptyResults.
@Test
@SuppressWarnings("unchecked")
public void testTopNCachingEmptyResults() throws Exception {
final TopNQueryBuilder builder = new TopNQueryBuilder().dataSource(DATA_SOURCE).dimension(TOP_DIM).metric("imps").threshold(3).intervals(SEG_SPEC).filters(DIM_FILTER).granularity(GRANULARITY).aggregators(AGGS).postAggregators(POST_AGGS).context(CONTEXT);
QueryRunner runner = new FinalizeResultsQueryRunner(client, new TopNQueryQueryToolChest(new TopNQueryConfig(), QueryRunnerTestHelper.NoopIntervalChunkingQueryRunnerDecorator()));
testQueryCaching(runner, builder.build(), new Interval("2011-01-01/2011-01-02"), makeTopNResultsWithoutRename(), new Interval("2011-01-02/2011-01-03"), makeTopNResultsWithoutRename(), new Interval("2011-01-05/2011-01-10"), makeTopNResultsWithoutRename(new DateTime("2011-01-05"), "a", 50, 4994, "b", 50, 4993, "c", 50, 4992, new DateTime("2011-01-06"), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, new DateTime("2011-01-07"), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, new DateTime("2011-01-08"), "a", 50, 4988, "b", 50, 4987, "c", 50, 4986, new DateTime("2011-01-09"), "a", 50, 4985, "b", 50, 4984, "c", 50, 4983), new Interval("2011-01-05/2011-01-10"), makeTopNResultsWithoutRename(new DateTime("2011-01-05T01"), "a", 50, 4994, "b", 50, 4993, "c", 50, 4992, new DateTime("2011-01-06T01"), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, new DateTime("2011-01-07T01"), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, new DateTime("2011-01-08T01"), "a", 50, 4988, "b", 50, 4987, "c", 50, 4986, new DateTime("2011-01-09T01"), "a", 50, 4985, "b", 50, 4984, "c", 50, 4983));
HashMap<String, List> context = new HashMap<String, List>();
TestHelper.assertExpectedResults(makeRenamedTopNResults(new DateTime("2011-01-05"), "a", 50, 4994, "b", 50, 4993, "c", 50, 4992, new DateTime("2011-01-05T01"), "a", 50, 4994, "b", 50, 4993, "c", 50, 4992, new DateTime("2011-01-06"), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, new DateTime("2011-01-06T01"), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, new DateTime("2011-01-07"), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, new DateTime("2011-01-07T01"), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, new DateTime("2011-01-08"), "a", 50, 4988, "b", 50, 4987, "c", 50, 4986, new DateTime("2011-01-08T01"), "a", 50, 4988, "b", 50, 4987, "c", 50, 4986, new DateTime("2011-01-09"), "a", 50, 4985, "b", 50, 4984, "c", 50, 4983, new DateTime("2011-01-09T01"), "a", 50, 4985, "b", 50, 4984, "c", 50, 4983), runner.run(builder.intervals("2011-01-01/2011-01-10").metric("imps").aggregators(RENAMED_AGGS).postAggregators(DIFF_ORDER_POST_AGGS).build(), context));
}
use of io.druid.query.QueryRunner in project druid by druid-io.
the class CachingClusteredClientTest method testQueryCaching.
@SuppressWarnings("unchecked")
public void testQueryCaching(final QueryRunner runner, final int numTimesToQuery, boolean expectBySegment, final Query query, // does this assume query intervals must be ordered?
Object... args) {
final List<Interval> queryIntervals = Lists.newArrayListWithCapacity(args.length / 2);
final List<List<Iterable<Result<Object>>>> expectedResults = Lists.newArrayListWithCapacity(queryIntervals.size());
parseResults(queryIntervals, expectedResults, args);
for (int i = 0; i < queryIntervals.size(); ++i) {
List<Object> mocks = Lists.newArrayList();
mocks.add(serverView);
final Interval actualQueryInterval = new Interval(queryIntervals.get(0).getStart(), queryIntervals.get(i).getEnd());
final List<Map<DruidServer, ServerExpectations>> serverExpectationList = populateTimeline(queryIntervals, expectedResults, i, mocks);
List<Capture> queryCaptures = Lists.newArrayList();
final Map<DruidServer, ServerExpectations> finalExpectation = serverExpectationList.get(serverExpectationList.size() - 1);
for (Map.Entry<DruidServer, ServerExpectations> entry : finalExpectation.entrySet()) {
DruidServer server = entry.getKey();
ServerExpectations expectations = entry.getValue();
EasyMock.expect(serverView.getQueryRunner(server)).andReturn(expectations.getQueryRunner()).once();
final Capture<? extends Query> capture = new Capture();
final Capture<? extends Map> context = new Capture();
queryCaptures.add(capture);
QueryRunner queryable = expectations.getQueryRunner();
if (query instanceof TimeseriesQuery) {
List<String> segmentIds = Lists.newArrayList();
List<Interval> intervals = Lists.newArrayList();
List<Iterable<Result<TimeseriesResultValue>>> results = Lists.newArrayList();
for (ServerExpectation expectation : expectations) {
segmentIds.add(expectation.getSegmentId());
intervals.add(expectation.getInterval());
results.add(expectation.getResults());
}
EasyMock.expect(queryable.run(EasyMock.capture(capture), EasyMock.capture(context))).andReturn(toQueryableTimeseriesResults(expectBySegment, segmentIds, intervals, results)).once();
} else if (query instanceof TopNQuery) {
List<String> segmentIds = Lists.newArrayList();
List<Interval> intervals = Lists.newArrayList();
List<Iterable<Result<TopNResultValue>>> results = Lists.newArrayList();
for (ServerExpectation expectation : expectations) {
segmentIds.add(expectation.getSegmentId());
intervals.add(expectation.getInterval());
results.add(expectation.getResults());
}
EasyMock.expect(queryable.run(EasyMock.capture(capture), EasyMock.capture(context))).andReturn(toQueryableTopNResults(segmentIds, intervals, results)).once();
} else if (query instanceof SearchQuery) {
List<String> segmentIds = Lists.newArrayList();
List<Interval> intervals = Lists.newArrayList();
List<Iterable<Result<SearchResultValue>>> results = Lists.newArrayList();
for (ServerExpectation expectation : expectations) {
segmentIds.add(expectation.getSegmentId());
intervals.add(expectation.getInterval());
results.add(expectation.getResults());
}
EasyMock.expect(queryable.run(EasyMock.capture(capture), EasyMock.capture(context))).andReturn(toQueryableSearchResults(segmentIds, intervals, results)).once();
} else if (query instanceof SelectQuery) {
List<String> segmentIds = Lists.newArrayList();
List<Interval> intervals = Lists.newArrayList();
List<Iterable<Result<SelectResultValue>>> results = Lists.newArrayList();
for (ServerExpectation expectation : expectations) {
segmentIds.add(expectation.getSegmentId());
intervals.add(expectation.getInterval());
results.add(expectation.getResults());
}
EasyMock.expect(queryable.run(EasyMock.capture(capture), EasyMock.capture(context))).andReturn(toQueryableSelectResults(segmentIds, intervals, results)).once();
} else if (query instanceof GroupByQuery) {
List<String> segmentIds = Lists.newArrayList();
List<Interval> intervals = Lists.newArrayList();
List<Iterable<Row>> results = Lists.newArrayList();
for (ServerExpectation expectation : expectations) {
segmentIds.add(expectation.getSegmentId());
intervals.add(expectation.getInterval());
results.add(expectation.getResults());
}
EasyMock.expect(queryable.run(EasyMock.capture(capture), EasyMock.capture(context))).andReturn(toQueryableGroupByResults(segmentIds, intervals, results)).once();
} else if (query instanceof TimeBoundaryQuery) {
List<String> segmentIds = Lists.newArrayList();
List<Interval> intervals = Lists.newArrayList();
List<Iterable<Result<TimeBoundaryResultValue>>> results = Lists.newArrayList();
for (ServerExpectation expectation : expectations) {
segmentIds.add(expectation.getSegmentId());
intervals.add(expectation.getInterval());
results.add(expectation.getResults());
}
EasyMock.expect(queryable.run(EasyMock.capture(capture), EasyMock.capture(context))).andReturn(toQueryableTimeBoundaryResults(segmentIds, intervals, results)).once();
} else {
throw new ISE("Unknown query type[%s]", query.getClass());
}
}
final int expectedResultsRangeStart;
final int expectedResultsRangeEnd;
if (query instanceof TimeBoundaryQuery) {
expectedResultsRangeStart = i;
expectedResultsRangeEnd = i + 1;
} else {
expectedResultsRangeStart = 0;
expectedResultsRangeEnd = i + 1;
}
runWithMocks(new Runnable() {
@Override
public void run() {
HashMap<String, List> context = new HashMap<String, List>();
for (int i = 0; i < numTimesToQuery; ++i) {
TestHelper.assertExpectedResults(new MergeIterable<>(Ordering.<Result<Object>>natural().nullsFirst(), FunctionalIterable.create(new RangeIterable(expectedResultsRangeStart, expectedResultsRangeEnd)).transformCat(new Function<Integer, Iterable<Iterable<Result<Object>>>>() {
@Override
public Iterable<Iterable<Result<Object>>> apply(@Nullable Integer input) {
List<Iterable<Result<Object>>> retVal = Lists.newArrayList();
final Map<DruidServer, ServerExpectations> exps = serverExpectationList.get(input);
for (ServerExpectations expectations : exps.values()) {
for (ServerExpectation expectation : expectations) {
retVal.add(expectation.getResults());
}
}
return retVal;
}
})), runner.run(query.withQuerySegmentSpec(new MultipleIntervalSegmentSpec(ImmutableList.of(actualQueryInterval))), context));
if (queryCompletedCallback != null) {
queryCompletedCallback.run();
}
}
}
}, mocks.toArray());
// make sure all the queries were sent down as 'bySegment'
for (Capture queryCapture : queryCaptures) {
Query capturedQuery = (Query) queryCapture.getValue();
if (expectBySegment) {
Assert.assertEquals(true, capturedQuery.getContextValue("bySegment"));
} else {
Assert.assertTrue(capturedQuery.getContextValue("bySegment") == null || capturedQuery.getContextValue("bySegment").equals(false));
}
}
}
}
Aggregations