From dccb59db94e502e43ffe77794ea672975e686182 Mon Sep 17 00:00:00 2001 From: Pankaj Kumar Date: Wed, 16 Aug 2017 23:40:42 +0800 Subject: [PATCH] HBASE-17617, Backport HBASE-16731 (Inconsistent results from the Get/Scan if we use the empty FilterList) to branch-1 --- .../java/org/apache/hadoop/hbase/client/Get.java | 5 + .../java/org/apache/hadoop/hbase/client/Query.java | 35 ++++ .../java/org/apache/hadoop/hbase/client/Scan.java | 36 +--- .../apache/hadoop/hbase/protobuf/ProtobufUtil.java | 9 +- .../org/apache/hadoop/hbase/client/TestScan.java | 32 ++++ .../hbase/protobuf/generated/ClientProtos.java | 181 +++++++++++++++++---- hbase-protocol/src/main/protobuf/Client.proto | 1 + .../apache/hadoop/hbase/regionserver/HRegion.java | 3 + .../hadoop/hbase/client/TestFromClientSide.java | 45 +++++ .../hadoop/hbase/protobuf/TestProtobufUtil.java | 1 + 10 files changed, 284 insertions(+), 64 deletions(-) diff --git a/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Get.java b/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Get.java index 88da0b0..72ab0ed 100644 --- a/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Get.java +++ b/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Get.java @@ -110,6 +110,7 @@ public class Get extends Query this.storeOffset = get.getRowOffsetPerColumnFamily(); this.tr = get.getTimeRange(); this.checkExistenceOnly = get.isCheckExistenceOnly(); + this.loadColumnFamiliesOnDemand = get.getLoadColumnFamiliesOnDemandValue(); this.closestRowBefore = get.isClosestRowBefore(); Map> fams = get.getFamilyMap(); for (Map.Entry> entry : fams.entrySet()) { @@ -241,6 +242,10 @@ public class Get extends Query return this; } + public Get setLoadColumnFamiliesOnDemand(boolean value) { + return (Get) super.setLoadColumnFamiliesOnDemand(value); + } + /** * Set the maximum number of values to return per row per Column Family * @param limit the maximum number of values returned / row / CF diff --git a/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Query.java b/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Query.java index 99d5a6a..7d3b628 100644 --- a/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Query.java +++ b/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Query.java @@ -42,6 +42,7 @@ public abstract class Query extends OperationWithAttributes { protected int targetReplicaId = -1; protected Consistency consistency = Consistency.STRONG; protected Map colFamTimeRangeMap = Maps.newTreeMap(Bytes.BYTES_COMPARATOR); + protected Boolean loadColumnFamiliesOnDemand = null; /** * @return Filter @@ -177,6 +178,40 @@ public abstract class Query extends OperationWithAttributes { IsolationLevel.fromBytes(attr); } + /** + * Set the value indicating whether loading CFs on demand should be allowed (cluster + * default is false). On-demand CF loading doesn't load column families until necessary, e.g. + * if you filter on one column, the other column family data will be loaded only for the rows + * that are included in result, not all rows like in normal case. + * With column-specific filters, like SingleColumnValueFilter w/filterIfMissing == true, + * this can deliver huge perf gains when there's a cf with lots of data; however, it can + * also lead to some inconsistent results, as follows: + * - if someone does a concurrent update to both column families in question you may get a row + * that never existed, e.g. for { rowKey = 5, { cat_videos => 1 }, { video => "my cat" } } + * someone puts rowKey 5 with { cat_videos => 0 }, { video => "my dog" }, concurrent scan + * filtering on "cat_videos == 1" can get { rowKey = 5, { cat_videos => 1 }, + * { video => "my dog" } }. + * - if there's a concurrent split and you have more than 2 column families, some rows may be + * missing some column families. + */ + public Query setLoadColumnFamiliesOnDemand(boolean value) { + this.loadColumnFamiliesOnDemand = value; + return this; + } + + /** + * Get the raw loadColumnFamiliesOnDemand setting; if it's not set, can be null. + */ + public Boolean getLoadColumnFamiliesOnDemandValue() { + return this.loadColumnFamiliesOnDemand; + } + + /** + * Get the logical value indicating whether on-demand CF loading should be allowed. + */ + public boolean doLoadColumnFamiliesOnDemand() { + return (this.loadColumnFamiliesOnDemand != null) && this.loadColumnFamiliesOnDemand; + } /** * Get versions of columns only within the specified timestamp range, diff --git a/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Scan.java b/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Scan.java index 6fcfe05..2893fc7 100644 --- a/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Scan.java +++ b/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Scan.java @@ -146,7 +146,6 @@ public class Scan extends Query { private TimeRange tr = new TimeRange(); private Map> familyMap = new TreeMap>(Bytes.BYTES_COMPARATOR); - private Boolean loadColumnFamiliesOnDemand = null; /** * Set it true for small scan to get better performance @@ -265,6 +264,7 @@ public class Scan extends Query { this.getScan = true; this.consistency = get.getConsistency(); this.setIsolationLevel(get.getIsolationLevel()); + this.loadColumnFamiliesOnDemand = get.getLoadColumnFamiliesOnDemandValue(); for (Map.Entry attr : get.getAttributesMap().entrySet()) { setAttribute(attr.getKey(), attr.getValue()); } @@ -725,40 +725,8 @@ public class Scan extends Query { return allowPartialResults; } - /** - * Set the value indicating whether loading CFs on demand should be allowed (cluster - * default is false). On-demand CF loading doesn't load column families until necessary, e.g. - * if you filter on one column, the other column family data will be loaded only for the rows - * that are included in result, not all rows like in normal case. - * With column-specific filters, like SingleColumnValueFilter w/filterIfMissing == true, - * this can deliver huge perf gains when there's a cf with lots of data; however, it can - * also lead to some inconsistent results, as follows: - * - if someone does a concurrent update to both column families in question you may get a row - * that never existed, e.g. for { rowKey = 5, { cat_videos => 1 }, { video => "my cat" } } - * someone puts rowKey 5 with { cat_videos => 0 }, { video => "my dog" }, concurrent scan - * filtering on "cat_videos == 1" can get { rowKey = 5, { cat_videos => 1 }, - * { video => "my dog" } }. - * - if there's a concurrent split and you have more than 2 column families, some rows may be - * missing some column families. - */ public Scan setLoadColumnFamiliesOnDemand(boolean value) { - this.loadColumnFamiliesOnDemand = value; - return this; - } - - /** - * Get the raw loadColumnFamiliesOnDemand setting; if it's not set, can be null. - */ - public Boolean getLoadColumnFamiliesOnDemandValue() { - return this.loadColumnFamiliesOnDemand; - } - - /** - * Get the logical value indicating whether on-demand CF loading should be allowed. - */ - public boolean doLoadColumnFamiliesOnDemand() { - return (this.loadColumnFamiliesOnDemand != null) - && this.loadColumnFamiliesOnDemand.booleanValue(); + return (Scan) super.setLoadColumnFamiliesOnDemand(value); } /** diff --git a/hbase-client/src/main/java/org/apache/hadoop/hbase/protobuf/ProtobufUtil.java b/hbase-client/src/main/java/org/apache/hadoop/hbase/protobuf/ProtobufUtil.java index 5461bd3..ed13617 100644 --- a/hbase-client/src/main/java/org/apache/hadoop/hbase/protobuf/ProtobufUtil.java +++ b/hbase-client/src/main/java/org/apache/hadoop/hbase/protobuf/ProtobufUtil.java @@ -512,6 +512,9 @@ public final class ProtobufUtil { if (proto.hasConsistency()) { get.setConsistency(toConsistency(proto.getConsistency())); } + if (proto.hasLoadColumnFamiliesOnDemand()) { + get.setLoadColumnFamiliesOnDemand(proto.getLoadColumnFamiliesOnDemand()); + } return get; } @@ -934,7 +937,7 @@ public final class ProtobufUtil { } Boolean loadColumnFamiliesOnDemand = scan.getLoadColumnFamiliesOnDemandValue(); if (loadColumnFamiliesOnDemand != null) { - scanBuilder.setLoadColumnFamiliesOnDemand(loadColumnFamiliesOnDemand.booleanValue()); + scanBuilder.setLoadColumnFamiliesOnDemand(loadColumnFamiliesOnDemand); } scanBuilder.setMaxVersions(scan.getMaxVersions()); for (Entry cftr : scan.getColumnFamilyTimeRange().entrySet()) { @@ -1161,6 +1164,10 @@ public final class ProtobufUtil { if (get.getConsistency() != null && get.getConsistency() != Consistency.STRONG) { builder.setConsistency(toConsistency(get.getConsistency())); } + Boolean loadColumnFamiliesOnDemand = get.getLoadColumnFamiliesOnDemandValue(); + if (loadColumnFamiliesOnDemand != null) { + builder.setLoadColumnFamiliesOnDemand(loadColumnFamiliesOnDemand); + } return builder.build(); } diff --git a/hbase-client/src/test/java/org/apache/hadoop/hbase/client/TestScan.java b/hbase-client/src/test/java/org/apache/hadoop/hbase/client/TestScan.java index d843723..c5b3a6b 100644 --- a/hbase-client/src/test/java/org/apache/hadoop/hbase/client/TestScan.java +++ b/hbase-client/src/test/java/org/apache/hadoop/hbase/client/TestScan.java @@ -19,6 +19,8 @@ package org.apache.hadoop.hbase.client; +import static org.junit.Assert.assertEquals; +import static org.junit.Assert.assertTrue; import static org.junit.Assert.fail; import java.io.IOException; @@ -26,6 +28,7 @@ import java.util.Arrays; import java.util.Set; import org.apache.hadoop.hbase.testclassification.SmallTests; +import org.apache.hadoop.hbase.filter.FilterList; import org.apache.hadoop.hbase.protobuf.ProtobufUtil; import org.apache.hadoop.hbase.protobuf.generated.ClientProtos; import org.apache.hadoop.hbase.security.visibility.Authorizations; @@ -56,6 +59,35 @@ public class TestScan { } @Test + public void testGetToScan() throws IOException { + Get get = new Get(Bytes.toBytes(1)); + get.setCacheBlocks(true).setConsistency(Consistency.TIMELINE).setFilter(new FilterList()) + .setId("get").setIsolationLevel(IsolationLevel.READ_COMMITTED) + .setLoadColumnFamiliesOnDemand(false).setMaxResultsPerColumnFamily(1000) + .setMaxVersions(9999).setRowOffsetPerColumnFamily(5).setTimeRange(0, 13) + .setAttribute("att_v0", Bytes.toBytes("att_v0")) + .setColumnFamilyTimeRange(Bytes.toBytes("cf"), 0, 123); + Scan scan = new Scan(get); + assertEquals(get.getCacheBlocks(), scan.getCacheBlocks()); + assertEquals(get.getConsistency(), scan.getConsistency()); + assertEquals(get.getFilter(), scan.getFilter()); + assertEquals(get.getId(), scan.getId()); + assertEquals(get.getIsolationLevel(), scan.getIsolationLevel()); + assertEquals(get.getLoadColumnFamiliesOnDemandValue(), + scan.getLoadColumnFamiliesOnDemandValue()); + assertEquals(get.getMaxResultsPerColumnFamily(), scan.getMaxResultsPerColumnFamily()); + assertEquals(get.getMaxVersions(), scan.getMaxVersions()); + assertEquals(get.getRowOffsetPerColumnFamily(), scan.getRowOffsetPerColumnFamily()); + assertEquals(get.getTimeRange().getMin(), scan.getTimeRange().getMin()); + assertEquals(get.getTimeRange().getMax(), scan.getTimeRange().getMax()); + assertTrue(Bytes.equals(get.getAttribute("att_v0"), scan.getAttribute("att_v0"))); + assertEquals(get.getColumnFamilyTimeRange().get(Bytes.toBytes("cf")).getMin(), + scan.getColumnFamilyTimeRange().get(Bytes.toBytes("cf")).getMin()); + assertEquals(get.getColumnFamilyTimeRange().get(Bytes.toBytes("cf")).getMax(), + scan.getColumnFamilyTimeRange().get(Bytes.toBytes("cf")).getMax()); + } + + @Test public void testScanAttributes() { Scan scan = new Scan(); Assert.assertTrue(scan.getAttributesMap().isEmpty()); diff --git a/hbase-protocol/src/main/java/org/apache/hadoop/hbase/protobuf/generated/ClientProtos.java b/hbase-protocol/src/main/java/org/apache/hadoop/hbase/protobuf/generated/ClientProtos.java index 1964140..ac41303 100644 --- a/hbase-protocol/src/main/java/org/apache/hadoop/hbase/protobuf/generated/ClientProtos.java +++ b/hbase-protocol/src/main/java/org/apache/hadoop/hbase/protobuf/generated/ClientProtos.java @@ -1980,6 +1980,24 @@ public final class ClientProtos { */ org.apache.hadoop.hbase.protobuf.generated.HBaseProtos.ColumnFamilyTimeRangeOrBuilder getCfTimeRangeOrBuilder( int index); + + // optional bool load_column_families_on_demand = 14; + /** + * optional bool load_column_families_on_demand = 14; + * + *
+     * DO NOT add defaults to load_column_families_on_demand. 
+     * 
+ */ + boolean hasLoadColumnFamiliesOnDemand(); + /** + * optional bool load_column_families_on_demand = 14; + * + *
+     * DO NOT add defaults to load_column_families_on_demand. 
+     * 
+ */ + boolean getLoadColumnFamiliesOnDemand(); } /** * Protobuf type {@code hbase.pb.Get} @@ -2136,6 +2154,11 @@ public final class ClientProtos { cfTimeRange_.add(input.readMessage(org.apache.hadoop.hbase.protobuf.generated.HBaseProtos.ColumnFamilyTimeRange.PARSER, extensionRegistry)); break; } + case 112: { + bitField0_ |= 0x00000400; + loadColumnFamiliesOnDemand_ = input.readBool(); + break; + } } } } catch (com.google.protobuf.InvalidProtocolBufferException e) { @@ -2485,6 +2508,30 @@ public final class ClientProtos { return cfTimeRange_.get(index); } + // optional bool load_column_families_on_demand = 14; + public static final int LOAD_COLUMN_FAMILIES_ON_DEMAND_FIELD_NUMBER = 14; + private boolean loadColumnFamiliesOnDemand_; + /** + * optional bool load_column_families_on_demand = 14; + * + *
+     * DO NOT add defaults to load_column_families_on_demand. 
+     * 
+ */ + public boolean hasLoadColumnFamiliesOnDemand() { + return ((bitField0_ & 0x00000400) == 0x00000400); + } + /** + * optional bool load_column_families_on_demand = 14; + * + *
+     * DO NOT add defaults to load_column_families_on_demand. 
+     * 
+ */ + public boolean getLoadColumnFamiliesOnDemand() { + return loadColumnFamiliesOnDemand_; + } + private void initFields() { row_ = com.google.protobuf.ByteString.EMPTY; column_ = java.util.Collections.emptyList(); @@ -2499,6 +2546,7 @@ public final class ClientProtos { closestRowBefore_ = false; consistency_ = org.apache.hadoop.hbase.protobuf.generated.ClientProtos.Consistency.STRONG; cfTimeRange_ = java.util.Collections.emptyList(); + loadColumnFamiliesOnDemand_ = false; } private byte memoizedIsInitialized = -1; public final boolean isInitialized() { @@ -2579,6 +2627,9 @@ public final class ClientProtos { for (int i = 0; i < cfTimeRange_.size(); i++) { output.writeMessage(13, cfTimeRange_.get(i)); } + if (((bitField0_ & 0x00000400) == 0x00000400)) { + output.writeBool(14, loadColumnFamiliesOnDemand_); + } getUnknownFields().writeTo(output); } @@ -2640,6 +2691,10 @@ public final class ClientProtos { size += com.google.protobuf.CodedOutputStream .computeMessageSize(13, cfTimeRange_.get(i)); } + if (((bitField0_ & 0x00000400) == 0x00000400)) { + size += com.google.protobuf.CodedOutputStream + .computeBoolSize(14, loadColumnFamiliesOnDemand_); + } size += getUnknownFields().getSerializedSize(); memoizedSerializedSize = size; return size; @@ -2719,6 +2774,11 @@ public final class ClientProtos { } result = result && getCfTimeRangeList() .equals(other.getCfTimeRangeList()); + result = result && (hasLoadColumnFamiliesOnDemand() == other.hasLoadColumnFamiliesOnDemand()); + if (hasLoadColumnFamiliesOnDemand()) { + result = result && (getLoadColumnFamiliesOnDemand() + == other.getLoadColumnFamiliesOnDemand()); + } result = result && getUnknownFields().equals(other.getUnknownFields()); return result; @@ -2784,6 +2844,10 @@ public final class ClientProtos { hash = (37 * hash) + CF_TIME_RANGE_FIELD_NUMBER; hash = (53 * hash) + getCfTimeRangeList().hashCode(); } + if (hasLoadColumnFamiliesOnDemand()) { + hash = (37 * hash) + LOAD_COLUMN_FAMILIES_ON_DEMAND_FIELD_NUMBER; + hash = (53 * hash) + hashBoolean(getLoadColumnFamiliesOnDemand()); + } hash = (29 * hash) + getUnknownFields().hashCode(); memoizedHashCode = hash; return hash; @@ -2952,6 +3016,8 @@ public final class ClientProtos { } else { cfTimeRangeBuilder_.clear(); } + loadColumnFamiliesOnDemand_ = false; + bitField0_ = (bitField0_ & ~0x00002000); return this; } @@ -3055,6 +3121,10 @@ public final class ClientProtos { } else { result.cfTimeRange_ = cfTimeRangeBuilder_.build(); } + if (((from_bitField0_ & 0x00002000) == 0x00002000)) { + to_bitField0_ |= 0x00000400; + } + result.loadColumnFamiliesOnDemand_ = loadColumnFamiliesOnDemand_; result.bitField0_ = to_bitField0_; onBuilt(); return result; @@ -3179,6 +3249,9 @@ public final class ClientProtos { } } } + if (other.hasLoadColumnFamiliesOnDemand()) { + setLoadColumnFamiliesOnDemand(other.getLoadColumnFamiliesOnDemand()); + } this.mergeUnknownFields(other.getUnknownFields()); return this; } @@ -4498,6 +4571,55 @@ public final class ClientProtos { return cfTimeRangeBuilder_; } + // optional bool load_column_families_on_demand = 14; + private boolean loadColumnFamiliesOnDemand_ ; + /** + * optional bool load_column_families_on_demand = 14; + * + *
+       * DO NOT add defaults to load_column_families_on_demand. 
+       * 
+ */ + public boolean hasLoadColumnFamiliesOnDemand() { + return ((bitField0_ & 0x00002000) == 0x00002000); + } + /** + * optional bool load_column_families_on_demand = 14; + * + *
+       * DO NOT add defaults to load_column_families_on_demand. 
+       * 
+ */ + public boolean getLoadColumnFamiliesOnDemand() { + return loadColumnFamiliesOnDemand_; + } + /** + * optional bool load_column_families_on_demand = 14; + * + *
+       * DO NOT add defaults to load_column_families_on_demand. 
+       * 
+ */ + public Builder setLoadColumnFamiliesOnDemand(boolean value) { + bitField0_ |= 0x00002000; + loadColumnFamiliesOnDemand_ = value; + onChanged(); + return this; + } + /** + * optional bool load_column_families_on_demand = 14; + * + *
+       * DO NOT add defaults to load_column_families_on_demand. 
+       * 
+ */ + public Builder clearLoadColumnFamiliesOnDemand() { + bitField0_ = (bitField0_ & ~0x00002000); + loadColumnFamiliesOnDemand_ = false; + onChanged(); + return this; + } + // @@protoc_insertion_point(builder_scope:hbase.pb.Get) } @@ -35622,7 +35744,7 @@ public final class ClientProtos { "o\032\017MapReduce.proto\"\037\n\016Authorizations\022\r\n\005" + "label\030\001 \003(\t\"$\n\016CellVisibility\022\022\n\nexpress" + "ion\030\001 \002(\t\"+\n\006Column\022\016\n\006family\030\001 \002(\014\022\021\n\tq" + - "ualifier\030\002 \003(\014\"\271\003\n\003Get\022\013\n\003row\030\001 \002(\014\022 \n\006c" + + "ualifier\030\002 \003(\014\"\341\003\n\003Get\022\013\n\003row\030\001 \002(\014\022 \n\006c" + "olumn\030\002 \003(\0132\020.hbase.pb.Column\022*\n\tattribu" + "te\030\003 \003(\0132\027.hbase.pb.NameBytesPair\022 \n\006fil" + "ter\030\004 \001(\0132\020.hbase.pb.Filter\022\'\n\ntime_rang" + @@ -35633,10 +35755,11 @@ public final class ClientProtos { "losest_row_before\030\013 \001(\010:\005false\0222\n\013consis" + "tency\030\014 \001(\0162\025.hbase.pb.Consistency:\006STRO" + "NG\0226\n\rcf_time_range\030\r \003(\0132\037.hbase.pb.Col" + - "umnFamilyTimeRange\"\203\001\n\006Result\022\034\n\004cell\030\001 " + + "umnFamilyTimeRange\022&\n\036load_column_famili" + + "es_on_demand\030\016 \001(\010\"\203\001\n\006Result\022\034\n\004cell\030\001 " + "\003(\0132\016.hbase.pb.Cell\022\035\n\025associated_cell_c" + - "ount\030\002 \001(\005\022\016\n\006exists\030\003 \001(\010\022\024\n\005stale\030\004 \001(" + - "\010:\005false\022\026\n\007partial\030\005 \001(\010:\005false\"S\n\nGetR", + "ount\030\002 \001(\005\022\016\n\006exists\030\003 \001(\010\022\024\n\005stale\030\004 \001(", + "\010:\005false\022\026\n\007partial\030\005 \001(\010:\005false\"S\n\nGetR" + "equest\022)\n\006region\030\001 \002(\0132\031.hbase.pb.Region" + "Specifier\022\032\n\003get\030\002 \002(\0132\r.hbase.pb.Get\"/\n" + "\013GetResponse\022 \n\006result\030\001 \001(\0132\020.hbase.pb." + @@ -35645,8 +35768,8 @@ public final class ClientProtos { "type\030\004 \002(\0162\025.hbase.pb.CompareType\022(\n\ncom" + "parator\030\005 \002(\0132\024.hbase.pb.Comparator\"\364\006\n\r" + "MutationProto\022\013\n\003row\030\001 \001(\014\0229\n\013mutate_typ" + - "e\030\002 \001(\0162$.hbase.pb.MutationProto.Mutatio" + - "nType\0229\n\014column_value\030\003 \003(\0132#.hbase.pb.M", + "e\030\002 \001(\0162$.hbase.pb.MutationProto.Mutatio", + "nType\0229\n\014column_value\030\003 \003(\0132#.hbase.pb.M" + "utationProto.ColumnValue\022\021\n\ttimestamp\030\004 " + "\001(\004\022*\n\tattribute\030\005 \003(\0132\027.hbase.pb.NameBy" + "tesPair\022C\n\ndurability\030\006 \001(\0162\".hbase.pb.M" + @@ -35655,8 +35778,8 @@ public final class ClientProtos { "\025associated_cell_count\030\010 \001(\005\022\r\n\005nonce\030\t " + "\001(\004\032\371\001\n\013ColumnValue\022\016\n\006family\030\001 \002(\014\022K\n\017q" + "ualifier_value\030\002 \003(\01322.hbase.pb.Mutation" + - "Proto.ColumnValue.QualifierValue\032\214\001\n\016Qua" + - "lifierValue\022\021\n\tqualifier\030\001 \001(\014\022\r\n\005value\030", + "Proto.ColumnValue.QualifierValue\032\214\001\n\016Qua", + "lifierValue\022\021\n\tqualifier\030\001 \001(\014\022\r\n\005value\030" + "\002 \001(\014\022\021\n\ttimestamp\030\003 \001(\004\0227\n\013delete_type\030" + "\004 \001(\0162\".hbase.pb.MutationProto.DeleteTyp" + "e\022\014\n\004tags\030\005 \001(\014\"W\n\nDurability\022\017\n\013USE_DEF" + @@ -35665,8 +35788,8 @@ public final class ClientProtos { "\022\n\n\006APPEND\020\000\022\r\n\tINCREMENT\020\001\022\007\n\003PUT\020\002\022\n\n\006" + "DELETE\020\003\"p\n\nDeleteType\022\026\n\022DELETE_ONE_VER" + "SION\020\000\022\034\n\030DELETE_MULTIPLE_VERSIONS\020\001\022\021\n\r" + - "DELETE_FAMILY\020\002\022\031\n\025DELETE_FAMILY_VERSION" + - "\020\003\"\242\001\n\rMutateRequest\022)\n\006region\030\001 \002(\0132\031.h", + "DELETE_FAMILY\020\002\022\031\n\025DELETE_FAMILY_VERSION", + "\020\003\"\242\001\n\rMutateRequest\022)\n\006region\030\001 \002(\0132\031.h" + "base.pb.RegionSpecifier\022)\n\010mutation\030\002 \002(" + "\0132\027.hbase.pb.MutationProto\022&\n\tcondition\030" + "\003 \001(\0132\023.hbase.pb.Condition\022\023\n\013nonce_grou" + @@ -35675,8 +35798,8 @@ public final class ClientProtos { "\004\n\004Scan\022 \n\006column\030\001 \003(\0132\020.hbase.pb.Colum" + "n\022*\n\tattribute\030\002 \003(\0132\027.hbase.pb.NameByte" + "sPair\022\021\n\tstart_row\030\003 \001(\014\022\020\n\010stop_row\030\004 \001" + - "(\014\022 \n\006filter\030\005 \001(\0132\020.hbase.pb.Filter\022\'\n\n" + - "time_range\030\006 \001(\0132\023.hbase.pb.TimeRange\022\027\n", + "(\014\022 \n\006filter\030\005 \001(\0132\020.hbase.pb.Filter\022\'\n\n", + "time_range\030\006 \001(\0132\023.hbase.pb.TimeRange\022\027\n" + "\014max_versions\030\007 \001(\r:\0011\022\032\n\014cache_blocks\030\010" + " \001(\010:\004true\022\022\n\nbatch_size\030\t \001(\r\022\027\n\017max_re" + "sult_size\030\n \001(\004\022\023\n\013store_limit\030\013 \001(\r\022\024\n\014" + @@ -35685,8 +35808,8 @@ public final class ClientProtos { "ersed\030\017 \001(\010:\005false\0222\n\013consistency\030\020 \001(\0162" + "\025.hbase.pb.Consistency:\006STRONG\022\017\n\007cachin" + "g\030\021 \001(\r\022\035\n\025allow_partial_results\030\022 \001(\010\0226" + - "\n\rcf_time_range\030\023 \003(\0132\037.hbase.pb.ColumnF" + - "amilyTimeRange\"\246\002\n\013ScanRequest\022)\n\006region", + "\n\rcf_time_range\030\023 \003(\0132\037.hbase.pb.ColumnF", + "amilyTimeRange\"\246\002\n\013ScanRequest\022)\n\006region" + "\030\001 \001(\0132\031.hbase.pb.RegionSpecifier\022\034\n\004sca" + "n\030\002 \001(\0132\016.hbase.pb.Scan\022\022\n\nscanner_id\030\003 " + "\001(\004\022\026\n\016number_of_rows\030\004 \001(\r\022\025\n\rclose_sca" + @@ -35695,8 +35818,8 @@ public final class ClientProtos { "dles_heartbeats\030\010 \001(\010\022\032\n\022track_scan_metr" + "ics\030\t \001(\010\022\024\n\005renew\030\n \001(\010:\005false\"\232\002\n\014Scan" + "Response\022\030\n\020cells_per_result\030\001 \003(\r\022\022\n\nsc" + - "anner_id\030\002 \001(\004\022\024\n\014more_results\030\003 \001(\010\022\013\n\003" + - "ttl\030\004 \001(\r\022!\n\007results\030\005 \003(\0132\020.hbase.pb.Re", + "anner_id\030\002 \001(\004\022\024\n\014more_results\030\003 \001(\010\022\013\n\003", + "ttl\030\004 \001(\r\022!\n\007results\030\005 \003(\0132\020.hbase.pb.Re" + "sult\022\r\n\005stale\030\006 \001(\010\022\037\n\027partial_flag_per_" + "result\030\007 \003(\010\022\036\n\026more_results_in_region\030\010" + " \001(\010\022\031\n\021heartbeat_message\030\t \001(\010\022+\n\014scan_" + @@ -35705,8 +35828,8 @@ public final class ClientProtos { "hbase.pb.RegionSpecifier\022>\n\013family_path\030" + "\002 \003(\0132).hbase.pb.BulkLoadHFileRequest.Fa" + "milyPath\022\026\n\016assign_seq_num\030\003 \001(\010\032*\n\nFami" + - "lyPath\022\016\n\006family\030\001 \002(\014\022\014\n\004path\030\002 \002(\t\"\'\n\025" + - "BulkLoadHFileResponse\022\016\n\006loaded\030\001 \002(\010\"a\n", + "lyPath\022\016\n\006family\030\001 \002(\014\022\014\n\004path\030\002 \002(\t\"\'\n\025", + "BulkLoadHFileResponse\022\016\n\006loaded\030\001 \002(\010\"a\n" + "\026CoprocessorServiceCall\022\013\n\003row\030\001 \002(\014\022\024\n\014" + "service_name\030\002 \002(\t\022\023\n\013method_name\030\003 \002(\t\022" + "\017\n\007request\030\004 \002(\014\"B\n\030CoprocessorServiceRe" + @@ -35715,8 +35838,8 @@ public final class ClientProtos { "ion\030\001 \002(\0132\031.hbase.pb.RegionSpecifier\022.\n\004" + "call\030\002 \002(\0132 .hbase.pb.CoprocessorService" + "Call\"o\n\032CoprocessorServiceResponse\022)\n\006re" + - "gion\030\001 \002(\0132\031.hbase.pb.RegionSpecifier\022&\n" + - "\005value\030\002 \002(\0132\027.hbase.pb.NameBytesPair\"\226\001", + "gion\030\001 \002(\0132\031.hbase.pb.RegionSpecifier\022&\n", + "\005value\030\002 \002(\0132\027.hbase.pb.NameBytesPair\"\226\001" + "\n\006Action\022\r\n\005index\030\001 \001(\r\022)\n\010mutation\030\002 \001(" + "\0132\027.hbase.pb.MutationProto\022\032\n\003get\030\003 \001(\0132" + "\r.hbase.pb.Get\0226\n\014service_call\030\004 \001(\0132 .h" + @@ -35725,8 +35848,8 @@ public final class ClientProtos { "Specifier\022\016\n\006atomic\030\002 \001(\010\022 \n\006action\030\003 \003(" + "\0132\020.hbase.pb.Action\"c\n\017RegionLoadStats\022\027" + "\n\014memstoreLoad\030\001 \001(\005:\0010\022\030\n\rheapOccupancy" + - "\030\002 \001(\005:\0010\022\035\n\022compactionPressure\030\003 \001(\005:\0010" + - "\"j\n\024MultiRegionLoadStats\022)\n\006region\030\001 \003(\013", + "\030\002 \001(\005:\0010\022\035\n\022compactionPressure\030\003 \001(\005:\0010", + "\"j\n\024MultiRegionLoadStats\022)\n\006region\030\001 \003(\013" + "2\031.hbase.pb.RegionSpecifier\022\'\n\004stat\030\002 \003(" + "\0132\031.hbase.pb.RegionLoadStats\"\336\001\n\021ResultO" + "rException\022\r\n\005index\030\001 \001(\r\022 \n\006result\030\002 \001(" + @@ -35735,8 +35858,8 @@ public final class ClientProtos { "t\030\004 \001(\0132\".hbase.pb.CoprocessorServiceRes" + "ult\0220\n\tloadStats\030\005 \001(\0132\031.hbase.pb.Region" + "LoadStatsB\002\030\001\"x\n\022RegionActionResult\0226\n\021r" + - "esultOrException\030\001 \003(\0132\033.hbase.pb.Result" + - "OrException\022*\n\texception\030\002 \001(\0132\027.hbase.p", + "esultOrException\030\001 \003(\0132\033.hbase.pb.Result", + "OrException\022*\n\texception\030\002 \001(\0132\027.hbase.p" + "b.NameBytesPair\"x\n\014MultiRequest\022,\n\014regio" + "nAction\030\001 \003(\0132\026.hbase.pb.RegionAction\022\022\n" + "\nnonceGroup\030\002 \001(\004\022&\n\tcondition\030\003 \001(\0132\023.h" + @@ -35745,8 +35868,8 @@ public final class ClientProtos { "nActionResult\022\021\n\tprocessed\030\002 \001(\010\0228\n\020regi" + "onStatistics\030\003 \001(\0132\036.hbase.pb.MultiRegio" + "nLoadStats*\'\n\013Consistency\022\n\n\006STRONG\020\000\022\014\n" + - "\010TIMELINE\020\0012\203\004\n\rClientService\0222\n\003Get\022\024.h" + - "base.pb.GetRequest\032\025.hbase.pb.GetRespons", + "\010TIMELINE\020\0012\203\004\n\rClientService\0222\n\003Get\022\024.h", + "base.pb.GetRequest\032\025.hbase.pb.GetRespons" + "e\022;\n\006Mutate\022\027.hbase.pb.MutateRequest\032\030.h" + "base.pb.MutateResponse\0225\n\004Scan\022\025.hbase.p" + "b.ScanRequest\032\026.hbase.pb.ScanResponse\022P\n" + @@ -35755,8 +35878,8 @@ public final class ClientProtos { "X\n\013ExecService\022#.hbase.pb.CoprocessorSer" + "viceRequest\032$.hbase.pb.CoprocessorServic" + "eResponse\022d\n\027ExecRegionServerService\022#.h" + - "base.pb.CoprocessorServiceRequest\032$.hbas" + - "e.pb.CoprocessorServiceResponse\0228\n\005Multi", + "base.pb.CoprocessorServiceRequest\032$.hbas", + "e.pb.CoprocessorServiceResponse\0228\n\005Multi" + "\022\026.hbase.pb.MultiRequest\032\027.hbase.pb.Mult" + "iResponseBB\n*org.apache.hadoop.hbase.pro" + "tobuf.generatedB\014ClientProtosH\001\210\001\001\240\001\001" @@ -35789,7 +35912,7 @@ public final class ClientProtos { internal_static_hbase_pb_Get_fieldAccessorTable = new com.google.protobuf.GeneratedMessage.FieldAccessorTable( internal_static_hbase_pb_Get_descriptor, - new java.lang.String[] { "Row", "Column", "Attribute", "Filter", "TimeRange", "MaxVersions", "CacheBlocks", "StoreLimit", "StoreOffset", "ExistenceOnly", "ClosestRowBefore", "Consistency", "CfTimeRange", }); + new java.lang.String[] { "Row", "Column", "Attribute", "Filter", "TimeRange", "MaxVersions", "CacheBlocks", "StoreLimit", "StoreOffset", "ExistenceOnly", "ClosestRowBefore", "Consistency", "CfTimeRange", "LoadColumnFamiliesOnDemand", }); internal_static_hbase_pb_Result_descriptor = getDescriptor().getMessageTypes().get(4); internal_static_hbase_pb_Result_fieldAccessorTable = new diff --git a/hbase-protocol/src/main/protobuf/Client.proto b/hbase-protocol/src/main/protobuf/Client.proto index 9cb1555..56f9f9f 100644 --- a/hbase-protocol/src/main/protobuf/Client.proto +++ b/hbase-protocol/src/main/protobuf/Client.proto @@ -88,6 +88,7 @@ message Get { optional Consistency consistency = 12 [default = STRONG]; repeated ColumnFamilyTimeRange cf_time_range = 13; + optional bool load_column_families_on_demand = 14; /* DO NOT add defaults to load_column_families_on_demand. */ } message Result { diff --git a/hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/HRegion.java b/hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/HRegion.java index 718fe74..906ea58 100644 --- a/hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/HRegion.java +++ b/hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/HRegion.java @@ -7080,6 +7080,9 @@ public class HRegion implements HeapSize, PropagatingConfigurationObserver, Regi scan = new Scan(get); } + if (scan.getLoadColumnFamiliesOnDemandValue() == null) { + scan.setLoadColumnFamiliesOnDemand(isLoadingCfsOnDemandDefault()); + } RegionScanner scanner = null; try { scanner = getScanner(scan, null, nonceGroup, nonce); diff --git a/hbase-server/src/test/java/org/apache/hadoop/hbase/client/TestFromClientSide.java b/hbase-server/src/test/java/org/apache/hadoop/hbase/client/TestFromClientSide.java index e7f614f..24ac4c6 100644 --- a/hbase-server/src/test/java/org/apache/hadoop/hbase/client/TestFromClientSide.java +++ b/hbase-server/src/test/java/org/apache/hadoop/hbase/client/TestFromClientSide.java @@ -34,6 +34,7 @@ import java.util.Collections; import java.util.HashMap; import java.util.HashSet; import java.util.Iterator; +import java.util.LinkedList; import java.util.List; import java.util.Map; import java.util.NavigableMap; @@ -5665,6 +5666,50 @@ public class TestFromClientSide { } @Test + public void testEmptyFilterList() throws Exception { + // Test Initialization. + TableName TABLE = TableName.valueOf("testEmptyFilterList"); + Table table = TEST_UTIL.createTable(TABLE, FAMILY); + + // Insert one row each region + Put put = new Put(Bytes.toBytes("row")); + put.addColumn(FAMILY, QUALIFIER, VALUE); + table.put(put); + + List scanResults = new LinkedList<>(); + Scan scan = new Scan(); + scan.setFilter(new FilterList()); + try (ResultScanner scanner = table.getScanner(scan)) { + for (Result r : scanner) { + scanResults.add(r); + } + } + + Get g = new Get(Bytes.toBytes("row")); + g.setFilter(new FilterList()); + Result getResult = table.get(g); + if (scanResults.isEmpty()) { + assertTrue(getResult.isEmpty()); + } else if (scanResults.size() == 1) { + Result scanResult = scanResults.get(0); + assertEquals(scanResult.rawCells().length, getResult.rawCells().length); + for (int i = 0; i != scanResult.rawCells().length; ++i) { + Cell scanCell = scanResult.rawCells()[i]; + Cell getCell = getResult.rawCells()[i]; + assertEquals(0, Bytes.compareTo(CellUtil.cloneRow(scanCell), CellUtil.cloneRow(getCell))); + assertEquals(0, + Bytes.compareTo(CellUtil.cloneFamily(scanCell), CellUtil.cloneFamily(getCell))); + assertEquals(0, + Bytes.compareTo(CellUtil.cloneQualifier(scanCell), CellUtil.cloneQualifier(getCell))); + assertEquals(0, + Bytes.compareTo(CellUtil.cloneValue(scanCell), CellUtil.cloneValue(getCell))); + } + } else { + fail("The result retrieved from SCAN and Get should be same"); + } + } + + @Test public void testSmallScan() throws Exception { // Test Initialization. TableName TABLE = TableName.valueOf("testSmallScan"); diff --git a/hbase-server/src/test/java/org/apache/hadoop/hbase/protobuf/TestProtobufUtil.java b/hbase-server/src/test/java/org/apache/hadoop/hbase/protobuf/TestProtobufUtil.java index ddbbb74..7cff7fd 100644 --- a/hbase-server/src/test/java/org/apache/hadoop/hbase/protobuf/TestProtobufUtil.java +++ b/hbase-server/src/test/java/org/apache/hadoop/hbase/protobuf/TestProtobufUtil.java @@ -82,6 +82,7 @@ public class TestProtobufUtil { columnBuilder.clear(); columnBuilder.setFamily(ByteString.copyFromUtf8("f2")); getBuilder.addColumn(columnBuilder.build()); + getBuilder.setLoadColumnFamiliesOnDemand(true); ClientProtos.Get proto = getBuilder.build(); // default fields -- 2.7.2.windows.1