|
| 1 | +/* |
| 2 | + * Copyright DataStax, Inc. |
| 3 | + * |
| 4 | + * Licensed under the Apache License, Version 2.0 (the "License"); |
| 5 | + * you may not use this file except in compliance with the License. |
| 6 | + * You may obtain a copy of the License at |
| 7 | + * |
| 8 | + * http://www.apache.org/licenses/LICENSE-2.0 |
| 9 | + * |
| 10 | + * Unless required by applicable law or agreed to in writing, software |
| 11 | + * distributed under the License is distributed on an "AS IS" BASIS, |
| 12 | + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 13 | + * See the License for the specific language governing permissions and |
| 14 | + * limitations under the License. |
| 15 | + */ |
| 16 | + |
| 17 | +package org.apache.cassandra.index.sai.functional; |
| 18 | + |
| 19 | +import java.math.BigDecimal; |
| 20 | +import java.math.BigInteger; |
| 21 | +import java.net.UnknownHostException; |
| 22 | +import java.nio.ByteBuffer; |
| 23 | +import java.util.Collection; |
| 24 | +import java.util.Date; |
| 25 | +import java.util.UUID; |
| 26 | +import java.util.stream.Collectors; |
| 27 | +import java.util.stream.Stream; |
| 28 | + |
| 29 | +import org.junit.Before; |
| 30 | +import org.junit.Test; |
| 31 | +import org.junit.runner.RunWith; |
| 32 | +import org.junit.runners.Parameterized; |
| 33 | +import org.slf4j.Logger; |
| 34 | +import org.slf4j.LoggerFactory; |
| 35 | + |
| 36 | +import org.apache.cassandra.index.sai.SAITester; |
| 37 | +import org.apache.cassandra.index.sai.SAIUtil; |
| 38 | +import org.apache.cassandra.index.sai.disk.format.Version; |
| 39 | +import org.apache.cassandra.locator.InetAddressAndPort; |
| 40 | + |
| 41 | +import static org.assertj.core.api.Assertions.assertThat; |
| 42 | + |
| 43 | +@RunWith(Parameterized.class) |
| 44 | +public class SaiDiskSizeTest extends SAITester |
| 45 | +{ |
| 46 | + private static final Logger logger = LoggerFactory.getLogger(SaiDiskSizeTest.class); |
| 47 | + |
| 48 | + @Parameterized.Parameter |
| 49 | + public Version version; |
| 50 | + |
| 51 | + @Parameterized.Parameter(1) |
| 52 | + public int size; |
| 53 | + |
| 54 | + @Parameterized.Parameter(2) |
| 55 | + public String pkSuffix; |
| 56 | + |
| 57 | + @Parameterized.Parameter(3) |
| 58 | + public int rowsPerPartition; |
| 59 | + |
| 60 | + @Parameterized.Parameters(name = "{0}, size={1}, pk={2}, partitionSize={3}") |
| 61 | + public static Collection<Object[]> data() |
| 62 | + { |
| 63 | + return Version.ALL.stream() |
| 64 | + .flatMap(v -> { |
| 65 | + switch (v.toString()) |
| 66 | + { |
| 67 | + case "aa": |
| 68 | + return Stream.of( |
| 69 | + new Object[]{ v, 13014, "pk", 1 }, |
| 70 | + new Object[]{ v, 13016, "pk, v_int", 2 }, |
| 71 | + new Object[]{ v, 14266, "pk, v_int", 100 }); |
| 72 | + case "ba": |
| 73 | + case "ca": |
| 74 | + case "db": |
| 75 | + case "dc": |
| 76 | + return Stream.of( |
| 77 | + new Object[]{ v, 65734, "pk", 1 }, |
| 78 | + new Object[]{ v, 57627, "pk, v_int", 2 }, |
| 79 | + new Object[]{ v, 28004, "pk, v_int", 100 }); |
| 80 | + case "eb": |
| 81 | + case "ec": |
| 82 | + return Stream.of( |
| 83 | + new Object[]{ v, 67370, "pk", 1 }, |
| 84 | + new Object[]{ v, 59263, "pk, v_int", 2 }, |
| 85 | + new Object[]{ v, 29640, "pk, v_int", 100 }); |
| 86 | + case "ed": |
| 87 | + return Stream.of( |
| 88 | + new Object[]{ v, 67378, "pk", 1 }, |
| 89 | + new Object[]{ v, 59271, "pk, v_int", 2 }, |
| 90 | + new Object[]{ v, 29648, "pk, v_int", 100 }); |
| 91 | + default: |
| 92 | + return // A new version assumes the latest size by default |
| 93 | + Stream.of( |
| 94 | + new Object[]{ v, 67378, "pk", 1 }, |
| 95 | + new Object[]{ v, 59271, "pk, v_int", 2 }, |
| 96 | + new Object[]{ v, 29648, "pk, v_int", 100 }); |
| 97 | + } |
| 98 | + }) |
| 99 | + .collect(Collectors.toList()); |
| 100 | + } |
| 101 | + |
| 102 | + @Before |
| 103 | + public void setVersion() |
| 104 | + { |
| 105 | + SAIUtil.setCurrentVersion(version); |
| 106 | + } |
| 107 | + |
| 108 | + @Test |
| 109 | + public void testIndexDiskSizeAcrossVersions() throws UnknownHostException |
| 110 | + { |
| 111 | + createTable("CREATE TABLE %s (" + |
| 112 | + "pk int, " + |
| 113 | + "v_ascii ascii, " + |
| 114 | + "v_bigint bigint, " + |
| 115 | + "v_blob blob, " + |
| 116 | + "v_boolean boolean, " + |
| 117 | + "v_decimal decimal, " + |
| 118 | + "v_double double, " + |
| 119 | + "v_float float, " + |
| 120 | + "v_int int, " + |
| 121 | + "v_text text, " + |
| 122 | + "v_timestamp timestamp, " + |
| 123 | + "v_uuid uuid, " + |
| 124 | + "v_varchar varchar, " + |
| 125 | + "v_varint varint, " + |
| 126 | + "v_timeuuid timeuuid, " + |
| 127 | + "v_inet inet, " + |
| 128 | + "v_date date, " + |
| 129 | + "v_time time, " + |
| 130 | + "v_smallint smallint, " + |
| 131 | + "v_tinyint tinyint, " + |
| 132 | + "v_duration duration, " + |
| 133 | + "PRIMARY KEY (" + pkSuffix + "))"); |
| 134 | + |
| 135 | + verifyNoIndexFiles(); |
| 136 | + createIndex("CREATE CUSTOM INDEX ON %s(v_int) USING 'StorageAttachedIndex'"); |
| 137 | + |
| 138 | + waitForTableIndexesQueryable(); |
| 139 | + |
| 140 | + // Split data into 2 sstable segments |
| 141 | + insertRows(0, 1000); |
| 142 | + flush(); |
| 143 | + insertRows(1000, 1000); |
| 144 | + flush(); |
| 145 | + |
| 146 | + long diskSize = indexDiskSpaceUse(); |
| 147 | + logger.info("SAI Version: {}, Index Disk Size: {} bytes", version, diskSize); |
| 148 | + assertThat(diskSize) |
| 149 | + .as("Disk size for SAI version %s", version) |
| 150 | + .isLessThanOrEqualTo(size); |
| 151 | + |
| 152 | + compact(); |
| 153 | + |
| 154 | + diskSize = indexDiskSpaceUse(); |
| 155 | + logger.info("SAI Version: {}, Index Disk Size: {} bytes", version, diskSize); |
| 156 | + assertThat(diskSize) |
| 157 | + .as("Disk size for SAI version %s", version) |
| 158 | + .isLessThanOrEqualTo(size); |
| 159 | + } |
| 160 | + |
| 161 | + private void insertRows(int size, int start) throws UnknownHostException |
| 162 | + { |
| 163 | + for (int i = start; i < start + size; i++) |
| 164 | + { |
| 165 | + execute("INSERT INTO %s (pk, v_ascii, v_bigint, v_blob, v_boolean, v_decimal, " + |
| 166 | + "v_double, v_float, v_int, v_text, v_timestamp, v_uuid, v_varchar, " + |
| 167 | + "v_varint, v_timeuuid, v_inet, v_date, v_time, v_smallint, v_tinyint, v_duration) " + |
| 168 | + "VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)", |
| 169 | + i % (size / rowsPerPartition), // Have 2 rows per partition |
| 170 | + "ascii_" + i, |
| 171 | + (long) i * 1000000, |
| 172 | + ByteBuffer.wrap(("blob_" + i).getBytes()), |
| 173 | + i % 2 == 0, |
| 174 | + new BigDecimal(i + ".123"), |
| 175 | + i * 1.5, |
| 176 | + (float) (i * 2.5), |
| 177 | + i * 2, |
| 178 | + "text_value_" + i, |
| 179 | + new Date(System.currentTimeMillis() + i * 1000L), |
| 180 | + UUID.randomUUID(), |
| 181 | + "varchar_" + i, |
| 182 | + BigInteger.valueOf(i).multiply(BigInteger.valueOf(100)), |
| 183 | + UUID.fromString("00000000-0000-1000-8000-" + |
| 184 | + String.format("%012d", i)), |
| 185 | + InetAddressAndPort.getByName("127.0.0." + (i % 256)).address, |
| 186 | + i + 1, |
| 187 | + (long) i * 1000000000L, |
| 188 | + (short) (i % 32767), |
| 189 | + (byte) (i % 128), |
| 190 | + org.apache.cassandra.cql3.Duration.newInstance(i % 12, i % 30, i * 1000000000L) |
| 191 | + ); |
| 192 | + } |
| 193 | + } |
| 194 | +} |
0 commit comments