Search in sources :

Example 11 with ColumnStatistics

use of org.apache.hadoop.hive.metastore.api.ColumnStatistics in project hive by apache.

the class TestHBaseAggregateStatsExtrapolation method MiddleOfPartitionsHaveBitVectorStatus.

@Test
public void MiddleOfPartitionsHaveBitVectorStatus() throws Exception {
    String dbName = "default";
    String tableName = "snp";
    long now = System.currentTimeMillis();
    List<FieldSchema> cols = new ArrayList<>();
    cols.add(new FieldSchema("col4", "long", "nocomment"));
    SerDeInfo serde = new SerDeInfo("serde", "seriallib", null);
    StorageDescriptor sd = new StorageDescriptor(cols, "file:/tmp", "input", "output", false, 0, serde, null, null, Collections.<String, String>emptyMap());
    List<FieldSchema> partCols = new ArrayList<>();
    partCols.add(new FieldSchema("ds", "string", ""));
    Table table = new Table(tableName, dbName, "me", (int) now, (int) now, 0, sd, partCols, Collections.<String, String>emptyMap(), null, null, null);
    store.createTable(table);
    List<List<String>> partVals = new ArrayList<>();
    for (int i = 0; i < 10; i++) {
        List<String> partVal = Arrays.asList("" + i);
        partVals.add(partVal);
        StorageDescriptor psd = new StorageDescriptor(sd);
        psd.setLocation("file:/tmp/default/hit/ds=" + partVal);
        Partition part = new Partition(partVal, dbName, tableName, (int) now, (int) now, psd, Collections.<String, String>emptyMap());
        store.addPartition(part);
        if (i > 2 && i < 7) {
            ColumnStatistics cs = new ColumnStatistics();
            ColumnStatisticsDesc desc = new ColumnStatisticsDesc(false, dbName, tableName);
            desc.setLastAnalyzed(now);
            desc.setPartName("ds=" + partVal);
            cs.setStatsDesc(desc);
            ColumnStatisticsObj obj = new ColumnStatisticsObj();
            obj.setColName("col4");
            obj.setColType("long");
            ColumnStatisticsData data = new ColumnStatisticsData();
            LongColumnStatsData dcsd = new LongColumnStatsData();
            dcsd.setHighValue(1000 + i);
            dcsd.setLowValue(-1000 - i);
            dcsd.setNumNulls(i);
            dcsd.setNumDVs(10 * i);
            dcsd.setBitVectors(bitVectors);
            data.setLongStats(dcsd);
            obj.setStatsData(data);
            cs.addToStatsObj(obj);
            store.updatePartitionColumnStatistics(cs, partVal);
        }
    }
    Checker statChecker = new Checker() {

        @Override
        public void checkStats(AggrStats aggrStats) throws Exception {
            Assert.assertEquals(4, aggrStats.getPartsFound());
            Assert.assertEquals(1, aggrStats.getColStatsSize());
            ColumnStatisticsObj cso = aggrStats.getColStats().get(0);
            Assert.assertEquals("col4", cso.getColName());
            Assert.assertEquals("long", cso.getColType());
            LongColumnStatsData lcsd = cso.getStatsData().getLongStats();
            Assert.assertEquals(1006, lcsd.getHighValue(), 0.01);
            Assert.assertEquals(-1006, lcsd.getLowValue(), 0.01);
            Assert.assertEquals(45, lcsd.getNumNulls());
            Assert.assertEquals(3, lcsd.getNumDVs());
        }
    };
    List<String> partNames = new ArrayList<>();
    for (int i = 0; i < 10; i++) {
        partNames.add("ds=" + i);
    }
    AggrStats aggrStats = store.get_aggr_stats_for(dbName, tableName, partNames, Arrays.asList("col4"));
    statChecker.checkStats(aggrStats);
}
Also used : ColumnStatistics(org.apache.hadoop.hive.metastore.api.ColumnStatistics) Partition(org.apache.hadoop.hive.metastore.api.Partition) Table(org.apache.hadoop.hive.metastore.api.Table) AggrStats(org.apache.hadoop.hive.metastore.api.AggrStats) FieldSchema(org.apache.hadoop.hive.metastore.api.FieldSchema) SerDeInfo(org.apache.hadoop.hive.metastore.api.SerDeInfo) ArrayList(java.util.ArrayList) StorageDescriptor(org.apache.hadoop.hive.metastore.api.StorageDescriptor) LongColumnStatsData(org.apache.hadoop.hive.metastore.api.LongColumnStatsData) ColumnStatisticsObj(org.apache.hadoop.hive.metastore.api.ColumnStatisticsObj) ColumnStatisticsDesc(org.apache.hadoop.hive.metastore.api.ColumnStatisticsDesc) ArrayList(java.util.ArrayList) List(java.util.List) ColumnStatisticsData(org.apache.hadoop.hive.metastore.api.ColumnStatisticsData) Test(org.junit.Test)

Example 12 with ColumnStatistics

use of org.apache.hadoop.hive.metastore.api.ColumnStatistics in project hive by apache.

the class TestHBaseAggregateStatsExtrapolation method allPartitionsHaveBitVectorStatusDouble.

@Test
public void allPartitionsHaveBitVectorStatusDouble() throws Exception {
    String dbName = "default";
    String tableName = "snp";
    long now = System.currentTimeMillis();
    List<FieldSchema> cols = new ArrayList<>();
    cols.add(new FieldSchema("col1_double", "double", "nocomment"));
    SerDeInfo serde = new SerDeInfo("serde", "seriallib", null);
    StorageDescriptor sd = new StorageDescriptor(cols, "file:/tmp", "input", "output", false, 0, serde, null, null, Collections.<String, String>emptyMap());
    List<FieldSchema> partCols = new ArrayList<>();
    partCols.add(new FieldSchema("ds", "string", ""));
    Table table = new Table(tableName, dbName, "me", (int) now, (int) now, 0, sd, partCols, Collections.<String, String>emptyMap(), null, null, null);
    store.createTable(table);
    List<List<String>> partVals = new ArrayList<>();
    for (int i = 0; i < 10; i++) {
        List<String> partVal = Arrays.asList("" + i);
        partVals.add(partVal);
        StorageDescriptor psd = new StorageDescriptor(sd);
        psd.setLocation("file:/tmp/default/hit/ds=" + partVal);
        Partition part = new Partition(partVal, dbName, tableName, (int) now, (int) now, psd, Collections.<String, String>emptyMap());
        store.addPartition(part);
        ColumnStatistics cs = new ColumnStatistics();
        ColumnStatisticsDesc desc = new ColumnStatisticsDesc(false, dbName, tableName);
        desc.setLastAnalyzed(now);
        desc.setPartName("ds=" + partVal);
        cs.setStatsDesc(desc);
        ColumnStatisticsObj obj = new ColumnStatisticsObj();
        obj.setColName("col1_double");
        obj.setColType("double");
        ColumnStatisticsData data = new ColumnStatisticsData();
        DoubleColumnStatsData dcsd = new DoubleColumnStatsData();
        dcsd.setHighValue(1000 + i);
        dcsd.setLowValue(-1000 - i);
        dcsd.setNumNulls(i);
        dcsd.setNumDVs(10 * i + 1);
        dcsd.setBitVectors(bitVectors);
        data.setDoubleStats(dcsd);
        obj.setStatsData(data);
        cs.addToStatsObj(obj);
        store.updatePartitionColumnStatistics(cs, partVal);
    }
    Checker statChecker = new Checker() {

        @Override
        public void checkStats(AggrStats aggrStats) throws Exception {
            Assert.assertEquals(10, aggrStats.getPartsFound());
            Assert.assertEquals(1, aggrStats.getColStatsSize());
            ColumnStatisticsObj cso = aggrStats.getColStats().get(0);
            Assert.assertEquals("col1_double", cso.getColName());
            Assert.assertEquals("double", cso.getColType());
            DoubleColumnStatsData lcsd = cso.getStatsData().getDoubleStats();
            Assert.assertEquals(1009, lcsd.getHighValue(), 0.01);
            Assert.assertEquals(-1009, lcsd.getLowValue(), 0.01);
            Assert.assertEquals(45, lcsd.getNumNulls());
            Assert.assertEquals(3, lcsd.getNumDVs());
        }
    };
    List<String> partNames = new ArrayList<>();
    for (int i = 0; i < 10; i++) {
        partNames.add("ds=" + i);
    }
    AggrStats aggrStats = store.get_aggr_stats_for(dbName, tableName, partNames, Arrays.asList("col1_double"));
    statChecker.checkStats(aggrStats);
}
Also used : ColumnStatistics(org.apache.hadoop.hive.metastore.api.ColumnStatistics) Partition(org.apache.hadoop.hive.metastore.api.Partition) Table(org.apache.hadoop.hive.metastore.api.Table) AggrStats(org.apache.hadoop.hive.metastore.api.AggrStats) FieldSchema(org.apache.hadoop.hive.metastore.api.FieldSchema) SerDeInfo(org.apache.hadoop.hive.metastore.api.SerDeInfo) ArrayList(java.util.ArrayList) StorageDescriptor(org.apache.hadoop.hive.metastore.api.StorageDescriptor) ColumnStatisticsObj(org.apache.hadoop.hive.metastore.api.ColumnStatisticsObj) DoubleColumnStatsData(org.apache.hadoop.hive.metastore.api.DoubleColumnStatsData) ColumnStatisticsDesc(org.apache.hadoop.hive.metastore.api.ColumnStatisticsDesc) ArrayList(java.util.ArrayList) List(java.util.List) ColumnStatisticsData(org.apache.hadoop.hive.metastore.api.ColumnStatisticsData) Test(org.junit.Test)

Example 13 with ColumnStatistics

use of org.apache.hadoop.hive.metastore.api.ColumnStatistics in project hive by apache.

the class TestHBaseAggregateStatsExtrapolation method allPartitionsHaveBitVectorStatusString.

@Test
public void allPartitionsHaveBitVectorStatusString() throws Exception {
    String dbName = "default";
    String tableName = "snp";
    long now = System.currentTimeMillis();
    List<FieldSchema> cols = new ArrayList<>();
    cols.add(new FieldSchema("col1_string", "string", "nocomment"));
    SerDeInfo serde = new SerDeInfo("serde", "seriallib", null);
    StorageDescriptor sd = new StorageDescriptor(cols, "file:/tmp", "input", "output", false, 0, serde, null, null, Collections.<String, String>emptyMap());
    List<FieldSchema> partCols = new ArrayList<>();
    partCols.add(new FieldSchema("ds", "string", ""));
    Table table = new Table(tableName, dbName, "me", (int) now, (int) now, 0, sd, partCols, Collections.<String, String>emptyMap(), null, null, null);
    store.createTable(table);
    List<List<String>> partVals = new ArrayList<>();
    for (int i = 0; i < 10; i++) {
        List<String> partVal = Arrays.asList("" + i);
        partVals.add(partVal);
        StorageDescriptor psd = new StorageDescriptor(sd);
        psd.setLocation("file:/tmp/default/hit/ds=" + partVal);
        Partition part = new Partition(partVal, dbName, tableName, (int) now, (int) now, psd, Collections.<String, String>emptyMap());
        store.addPartition(part);
        ColumnStatistics cs = new ColumnStatistics();
        ColumnStatisticsDesc desc = new ColumnStatisticsDesc(false, dbName, tableName);
        desc.setLastAnalyzed(now);
        desc.setPartName("ds=" + partVal);
        cs.setStatsDesc(desc);
        ColumnStatisticsObj obj = new ColumnStatisticsObj();
        obj.setColName("col1_string");
        obj.setColType("string");
        ColumnStatisticsData data = new ColumnStatisticsData();
        StringColumnStatsData dcsd = new StringColumnStatsData();
        dcsd.setAvgColLen(i + 1);
        dcsd.setMaxColLen(i + 10);
        dcsd.setNumNulls(i);
        dcsd.setNumDVs(10 * i + 1);
        dcsd.setBitVectors(bitVectors);
        data.setStringStats(dcsd);
        obj.setStatsData(data);
        cs.addToStatsObj(obj);
        store.updatePartitionColumnStatistics(cs, partVal);
    }
    Checker statChecker = new Checker() {

        @Override
        public void checkStats(AggrStats aggrStats) throws Exception {
            Assert.assertEquals(10, aggrStats.getPartsFound());
            Assert.assertEquals(1, aggrStats.getColStatsSize());
            ColumnStatisticsObj cso = aggrStats.getColStats().get(0);
            Assert.assertEquals("col1_string", cso.getColName());
            Assert.assertEquals("string", cso.getColType());
            StringColumnStatsData lcsd = cso.getStatsData().getStringStats();
            Assert.assertEquals(10, lcsd.getAvgColLen(), 0.01);
            Assert.assertEquals(19, lcsd.getMaxColLen(), 0.01);
            Assert.assertEquals(45, lcsd.getNumNulls());
            Assert.assertEquals(3, lcsd.getNumDVs());
        }
    };
    List<String> partNames = new ArrayList<>();
    for (int i = 0; i < 10; i++) {
        partNames.add("ds=" + i);
    }
    AggrStats aggrStats = store.get_aggr_stats_for(dbName, tableName, partNames, Arrays.asList("col1_string"));
    statChecker.checkStats(aggrStats);
}
Also used : ColumnStatistics(org.apache.hadoop.hive.metastore.api.ColumnStatistics) Partition(org.apache.hadoop.hive.metastore.api.Partition) Table(org.apache.hadoop.hive.metastore.api.Table) AggrStats(org.apache.hadoop.hive.metastore.api.AggrStats) FieldSchema(org.apache.hadoop.hive.metastore.api.FieldSchema) SerDeInfo(org.apache.hadoop.hive.metastore.api.SerDeInfo) ArrayList(java.util.ArrayList) StorageDescriptor(org.apache.hadoop.hive.metastore.api.StorageDescriptor) StringColumnStatsData(org.apache.hadoop.hive.metastore.api.StringColumnStatsData) ColumnStatisticsObj(org.apache.hadoop.hive.metastore.api.ColumnStatisticsObj) ColumnStatisticsDesc(org.apache.hadoop.hive.metastore.api.ColumnStatisticsDesc) ArrayList(java.util.ArrayList) List(java.util.List) ColumnStatisticsData(org.apache.hadoop.hive.metastore.api.ColumnStatisticsData) Test(org.junit.Test)

Example 14 with ColumnStatistics

use of org.apache.hadoop.hive.metastore.api.ColumnStatistics in project hive by apache.

the class TestHBaseAggregateStatsExtrapolation method TwoEndsAndMiddleOfPartitionsHaveBitVectorStatusLong.

@Test
public void TwoEndsAndMiddleOfPartitionsHaveBitVectorStatusLong() throws Exception {
    String dbName = "default";
    String tableName = "snp";
    long now = System.currentTimeMillis();
    List<FieldSchema> cols = new ArrayList<>();
    cols.add(new FieldSchema("col5", "long", "nocomment"));
    SerDeInfo serde = new SerDeInfo("serde", "seriallib", null);
    StorageDescriptor sd = new StorageDescriptor(cols, "file:/tmp", "input", "output", false, 0, serde, null, null, Collections.<String, String>emptyMap());
    List<FieldSchema> partCols = new ArrayList<>();
    partCols.add(new FieldSchema("ds", "string", ""));
    Table table = new Table(tableName, dbName, "me", (int) now, (int) now, 0, sd, partCols, Collections.<String, String>emptyMap(), null, null, null);
    store.createTable(table);
    List<List<String>> partVals = new ArrayList<>();
    for (int i = 0; i < 10; i++) {
        List<String> partVal = Arrays.asList("" + i);
        partVals.add(partVal);
        StorageDescriptor psd = new StorageDescriptor(sd);
        psd.setLocation("file:/tmp/default/hit/ds=" + partVal);
        Partition part = new Partition(partVal, dbName, tableName, (int) now, (int) now, psd, Collections.<String, String>emptyMap());
        store.addPartition(part);
        if (i == 0 || i == 2 || i == 3 || i == 5 || i == 6 || i == 8) {
            ColumnStatistics cs = new ColumnStatistics();
            ColumnStatisticsDesc desc = new ColumnStatisticsDesc(false, dbName, tableName);
            desc.setLastAnalyzed(now);
            desc.setPartName("ds=" + partVal);
            cs.setStatsDesc(desc);
            ColumnStatisticsObj obj = new ColumnStatisticsObj();
            obj.setColName("col5");
            obj.setColType("long");
            ColumnStatisticsData data = new ColumnStatisticsData();
            LongColumnStatsData dcsd = new LongColumnStatsData();
            dcsd.setHighValue(1000 + i);
            dcsd.setLowValue(-1000 - i);
            dcsd.setNumNulls(i);
            dcsd.setNumDVs(10 * i);
            dcsd.setBitVectors(bitVectors);
            data.setLongStats(dcsd);
            obj.setStatsData(data);
            cs.addToStatsObj(obj);
            store.updatePartitionColumnStatistics(cs, partVal);
        }
    }
    Checker statChecker = new Checker() {

        @Override
        public void checkStats(AggrStats aggrStats) throws Exception {
            Assert.assertEquals(6, aggrStats.getPartsFound());
            Assert.assertEquals(1, aggrStats.getColStatsSize());
            ColumnStatisticsObj cso = aggrStats.getColStats().get(0);
            Assert.assertEquals("col5", cso.getColName());
            Assert.assertEquals("long", cso.getColType());
            LongColumnStatsData lcsd = cso.getStatsData().getLongStats();
            Assert.assertEquals(1010, lcsd.getHighValue(), 0.01);
            Assert.assertEquals(-1010, lcsd.getLowValue(), 0.01);
            Assert.assertEquals(40, lcsd.getNumNulls());
            Assert.assertEquals(3, lcsd.getNumDVs());
        }
    };
    List<String> partNames = new ArrayList<>();
    for (int i = 0; i < 10; i++) {
        partNames.add("ds=" + i);
    }
    AggrStats aggrStats = store.get_aggr_stats_for(dbName, tableName, partNames, Arrays.asList("col5"));
    statChecker.checkStats(aggrStats);
}
Also used : ColumnStatistics(org.apache.hadoop.hive.metastore.api.ColumnStatistics) Partition(org.apache.hadoop.hive.metastore.api.Partition) Table(org.apache.hadoop.hive.metastore.api.Table) AggrStats(org.apache.hadoop.hive.metastore.api.AggrStats) FieldSchema(org.apache.hadoop.hive.metastore.api.FieldSchema) SerDeInfo(org.apache.hadoop.hive.metastore.api.SerDeInfo) ArrayList(java.util.ArrayList) StorageDescriptor(org.apache.hadoop.hive.metastore.api.StorageDescriptor) LongColumnStatsData(org.apache.hadoop.hive.metastore.api.LongColumnStatsData) ColumnStatisticsObj(org.apache.hadoop.hive.metastore.api.ColumnStatisticsObj) ColumnStatisticsDesc(org.apache.hadoop.hive.metastore.api.ColumnStatisticsDesc) ArrayList(java.util.ArrayList) List(java.util.List) ColumnStatisticsData(org.apache.hadoop.hive.metastore.api.ColumnStatisticsData) Test(org.junit.Test)

Example 15 with ColumnStatistics

use of org.apache.hadoop.hive.metastore.api.ColumnStatistics in project hive by apache.

the class TestHBaseStore method longPartitionStatistics.

@Test
public void longPartitionStatistics() throws Exception {
    createMockTableAndPartition(INT_TYPE, INT_VAL);
    // Add partition stats for: LONG_COL and partition: {PART_KEY, INT_VAL} to DB
    // Because of the way our mock implementation works we actually need to not create the table
    // before we set statistics on it.
    ColumnStatistics stats = new ColumnStatistics();
    // Get a default ColumnStatisticsDesc for partition level stats
    ColumnStatisticsDesc desc = getMockPartColStatsDesc(PART_KEY, INT_VAL);
    stats.setStatsDesc(desc);
    // Get one of the pre-created ColumnStatisticsObj
    ColumnStatisticsObj obj = longColStatsObjs.get(0);
    LongColumnStatsData longData = obj.getStatsData().getLongStats();
    // Add to DB
    stats.addToStatsObj(obj);
    List<String> parVals = new ArrayList<String>();
    parVals.add(INT_VAL);
    store.updatePartitionColumnStatistics(stats, parVals);
    // Get from DB
    List<String> partNames = new ArrayList<String>();
    partNames.add(desc.getPartName());
    List<String> colNames = new ArrayList<String>();
    colNames.add(obj.getColName());
    List<ColumnStatistics> statsFromDB = store.getPartitionColumnStatistics(DB, TBL, partNames, colNames);
    // Compare ColumnStatisticsDesc
    Assert.assertEquals(1, statsFromDB.size());
    Assert.assertEquals(desc.getLastAnalyzed(), statsFromDB.get(0).getStatsDesc().getLastAnalyzed());
    Assert.assertEquals(DB, statsFromDB.get(0).getStatsDesc().getDbName());
    Assert.assertEquals(TBL, statsFromDB.get(0).getStatsDesc().getTableName());
    Assert.assertFalse(statsFromDB.get(0).getStatsDesc().isIsTblLevel());
    // Compare ColumnStatisticsObj
    Assert.assertEquals(1, statsFromDB.get(0).getStatsObjSize());
    ColumnStatisticsObj objFromDB = statsFromDB.get(0).getStatsObj().get(0);
    ColumnStatisticsData dataFromDB = objFromDB.getStatsData();
    // Compare ColumnStatisticsData
    Assert.assertEquals(ColumnStatisticsData._Fields.LONG_STATS, dataFromDB.getSetField());
    // Compare LongColumnStatsData
    LongColumnStatsData longDataFromDB = dataFromDB.getLongStats();
    Assert.assertEquals(longData.getHighValue(), longDataFromDB.getHighValue());
    Assert.assertEquals(longData.getLowValue(), longDataFromDB.getLowValue());
    Assert.assertEquals(longData.getNumNulls(), longDataFromDB.getNumNulls());
    Assert.assertEquals(longData.getNumDVs(), longDataFromDB.getNumDVs());
}
Also used : ColumnStatistics(org.apache.hadoop.hive.metastore.api.ColumnStatistics) ColumnStatisticsObj(org.apache.hadoop.hive.metastore.api.ColumnStatisticsObj) ColumnStatisticsDesc(org.apache.hadoop.hive.metastore.api.ColumnStatisticsDesc) ArrayList(java.util.ArrayList) LongColumnStatsData(org.apache.hadoop.hive.metastore.api.LongColumnStatsData) ColumnStatisticsData(org.apache.hadoop.hive.metastore.api.ColumnStatisticsData) Test(org.junit.Test)

Aggregations

ColumnStatistics (org.apache.hadoop.hive.metastore.api.ColumnStatistics)74 ColumnStatisticsObj (org.apache.hadoop.hive.metastore.api.ColumnStatisticsObj)65 ColumnStatisticsDesc (org.apache.hadoop.hive.metastore.api.ColumnStatisticsDesc)58 ColumnStatisticsData (org.apache.hadoop.hive.metastore.api.ColumnStatisticsData)57 ArrayList (java.util.ArrayList)49 Test (org.junit.Test)48 FieldSchema (org.apache.hadoop.hive.metastore.api.FieldSchema)31 Table (org.apache.hadoop.hive.metastore.api.Table)29 StorageDescriptor (org.apache.hadoop.hive.metastore.api.StorageDescriptor)28 SerDeInfo (org.apache.hadoop.hive.metastore.api.SerDeInfo)27 Partition (org.apache.hadoop.hive.metastore.api.Partition)26 AggrStats (org.apache.hadoop.hive.metastore.api.AggrStats)24 List (java.util.List)22 LongColumnStatsData (org.apache.hadoop.hive.metastore.api.LongColumnStatsData)19 DoubleColumnStatsData (org.apache.hadoop.hive.metastore.api.DoubleColumnStatsData)13 StringColumnStatsData (org.apache.hadoop.hive.metastore.api.StringColumnStatsData)13 BooleanColumnStatsData (org.apache.hadoop.hive.metastore.api.BooleanColumnStatsData)12 DecimalColumnStatsData (org.apache.hadoop.hive.metastore.api.DecimalColumnStatsData)9 MetaException (org.apache.hadoop.hive.metastore.api.MetaException)9 HashMap (java.util.HashMap)6