maedhroz commented on code in PR #2961: URL: https://github.com/apache/cassandra/pull/2961#discussion_r1434438572
########## test/distributed/org/apache/cassandra/distributed/test/guardrails/GuardrailNonPartitionRestrictedQueryTest.java: ########## @@ -0,0 +1,450 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license agreements. See the NOTICE file + * distributed with this work for additional information + * regarding copyright ownership. The ASF licenses this file + * to you under the Apache License, Version 2.0 (the + * "License"); you may not use this file except in compliance + * with the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package org.apache.cassandra.distributed.test.guardrails; + +import java.io.IOException; +import java.util.List; +import java.util.concurrent.TimeUnit; +import java.util.stream.Stream; + +import com.google.common.util.concurrent.Uninterruptibles; +import org.junit.AfterClass; +import org.junit.Before; +import org.junit.BeforeClass; +import org.junit.Test; + +import com.datastax.driver.core.ResultSet; +import com.datastax.driver.core.Session; +import com.datastax.driver.core.SimpleStatement; +import com.datastax.driver.core.exceptions.InvalidQueryException; +import org.apache.cassandra.cql3.QueryProcessor; +import org.apache.cassandra.db.ColumnFamilyStore; +import org.apache.cassandra.db.Keyspace; +import org.apache.cassandra.db.guardrails.Guardrails; +import org.apache.cassandra.distributed.Cluster; +import org.apache.cassandra.distributed.api.ConsistencyLevel; +import org.apache.cassandra.distributed.api.Feature; +import org.apache.cassandra.distributed.api.IInvokableInstance; +import org.apache.cassandra.distributed.api.IIsolatedExecutor; +import org.apache.cassandra.distributed.util.Auth; +import org.apache.cassandra.exceptions.QueryReferencesTooManyIndexesAbortException; +import org.apache.cassandra.service.ClientWarn; +import org.apache.cassandra.service.QueryState; +import org.apache.cassandra.service.reads.thresholds.CoordinatorWarnings; + +import static com.google.common.collect.Iterables.getOnlyElement; +import static org.apache.cassandra.db.ConsistencyLevel.ALL; +import static org.apache.cassandra.service.reads.thresholds.WarningsSnapshot.tooManyIndexesReadWarnMessage; +import static org.assertj.core.api.Assertions.assertThat; +import static org.junit.Assert.assertEquals; +import static org.junit.Assert.assertFalse; +import static org.junit.Assert.assertTrue; +import static org.junit.Assert.fail; + +public class GuardrailNonPartitionRestrictedQueryTest extends GuardrailTester +{ + private static Cluster cluster; + private static com.datastax.driver.core.Cluster driverCluster; + private static Session driverSession; + + @BeforeClass + public static void setupCluster() throws IOException + { + cluster = Cluster.build(3) + .withConfig(c -> c.with(Feature.GOSSIP, Feature.NATIVE_PROTOCOL) + .set("read_thresholds_enabled", "true") + .set("authenticator", "PasswordAuthenticator")) + .withDataDirCount(1) + .start(); + + Auth.waitForExistingRoles(cluster.get(1)); + + // create a regular user, since the default superuser is excluded from guardrails + com.datastax.driver.core.Cluster.Builder builder = com.datastax.driver.core.Cluster.builder().addContactPoint("127.0.0.1"); + try (com.datastax.driver.core.Cluster c = builder.withCredentials("cassandra", "cassandra").build(); + Session session = c.connect()) + { + session.execute("CREATE USER test WITH PASSWORD 'test'"); + } + + // connect using that superuser, we use the driver to get access to the client warnings + driverCluster = builder.withCredentials("test", "test").build(); + driverSession = driverCluster.connect(); + } + + @AfterClass + public static void teardownCluster() + { + if (driverSession != null) + driverSession.close(); + + if (driverCluster != null) + driverCluster.close(); + + if (cluster != null) + cluster.close(); + } + + @Before + public void beforeTest() + { + super.beforeTest(); + cluster.schemaChange("DROP KEYSPACE IF EXISTS " + KEYSPACE); + init(cluster); + } + + @Override + protected Cluster getCluster() + { + return cluster; + } + + @Test + public void testGuardrailForLegacy2i() + { + prepareSchema(false); + testGuardrailInternal(); + } + + @Test + public void testGuardrailForSAI() + { + prepareSchema(true); + testGuardrailInternal(); + } + + @Test + public void testSAIWarnThreshold() + { + prepareSchema(true); + setThresholds(2, 5); + + // flushing just 1 table will not trigger any threshold + long valueToQuery = createSSTables(1); + assertThat(executeSelect(valueToQuery, false)).isNull(); + + assertWarnAborts(0, 0); + + // create 3 more SSTables on each node, this will trigger warn threshold (3 > 2 but < 5) + valueToQuery = createSSTables(3); + String expectedMessage = tooManyIndexesReadWarnMessage(cluster.size(), + 3, + String.format("SELECT * FROM %s.%s WHERE v1 = %s ALLOW FILTERING", + KEYSPACE, tableName, valueToQuery)); + assertThat(getOnlyElement(executeSelect(valueToQuery, false))).contains(expectedMessage); + + assertWarnAborts(1, 0); + + // we effectively stop reacting on warnings + setThresholds(32, -1); + assertThat(executeSelect(valueToQuery, false)).isNull(); + + assertWarnAborts(1, 0); + + // we compacted SSTables of node 1 but not on 2 and 3, + // we set tresholds to all nodes hence we violated thresholds on node 2 and 3 + compact(1); + setThresholds(2, 5); + + // notice we expect warnings from 2 nodes + expectedMessage = tooManyIndexesReadWarnMessage(cluster.size() - 1, + 3, + String.format("SELECT * FROM %s.%s WHERE v1 = %s ALLOW FILTERING", + KEYSPACE, tableName, valueToQuery)); + + assertThat(getOnlyElement(executeSelect(valueToQuery, false))).contains(expectedMessage); + + assertWarnAborts(2, 0); + + // disable warn thresholds on nodes 2 and 3 + setThresholds(32, -1, 2, 3); + // we will not warn because nodes 2 and 3 are disabled and 1 does not violate warn threshold + assertThat(executeSelect(valueToQuery, false)).isNull(); + + assertWarnAborts(2, 0); + + // set thresholds back on nodes 2 and 3 + setThresholds(2, 5, 2, 3); + assertThat(getOnlyElement(executeSelect(valueToQuery, false))).contains(expectedMessage); + + assertWarnAborts(3, 0); + + // we compacted SSTables on the third node as well, so we will not violate any thresholds + compact(2); + compact(3); + + assertThat(executeSelect(valueToQuery, false)).isNull(); + + // metrics stay same as they were + assertWarnAborts(3, 0); + } + + @Test + public void testSAIFailThreshold() + { + prepareSchema(true); + setThresholds(2, 5); + + // create 6 SSTables on each node, this will trigger fail threshold (6 > 5) + long valueToQuery = createSSTables(6); + + String expectedMessage = String.format("referenced %s SSTable indexes for a query without restrictions on partition key " + + "and aborted the query SELECT * FROM %s.%s WHERE v1 = %s ALLOW FILTERING", + 6, KEYSPACE, tableName, valueToQuery); + + assertThat(getOnlyElement(executeSelect(valueToQuery, true))).contains(expectedMessage); + + assertWarnAborts(0, 1); + + // we effectively stop reacting on failures + setThresholds(32, -1); + assertThat(executeSelect(valueToQuery, false)).isNull(); + + // we compacted SSTables of node 1 but not on 2 and 3, + // we set tresholds to all nodes hence we violated thresholds on node 2 and 3 + compact(1); + setThresholds(2, 5); + assertThat(getOnlyElement(executeSelect(valueToQuery, true))).contains(expectedMessage); + + assertWarnAborts(0, 2); + + // disable fail thresholds on nodes 2 and 3 + setThresholds(32, -1, 2, 3); + // we will not fail because nodes 2 and 3 are disabled and 1 does not violate fail threshold + assertThat(executeSelect(valueToQuery, false)).isNull(); + + // set thresholds back on nodes 2 and 3 + setThresholds(2, 5, 2, 3); + assertThat(getOnlyElement(executeSelect(valueToQuery, true))).contains(expectedMessage); + + assertWarnAborts(0, 3); + + // we compacted SSTables on the second and the third node as well, so we will not violate any thresholds + compact(2); + compact(3); + + assertThat(executeSelect(valueToQuery, false)).isNull(); + + // metrics stay same as they were + assertWarnAborts(0, 3); + } + + private void testGuardrailInternal() + { + enableGuardrail(); + assertThat(executeSelectViaDriver()).isEmpty(); + + disableGuardrail(); + + try + { + executeSelectViaDriver(); + fail("selection on non-partition restricted queries should be forbidden"); + } + catch (InvalidQueryException e) + { + assertThat(e.getMessage()).contains(Guardrails.nonPartitionRestrictedIndexQueryEnabled.reason); + } + + // even we disabled guardrail, if we query by primary key and clustering column, if passes + executeViaDriver(String.format("SELECT * from %s.%s WHERE k = 0 AND c = 0", KEYSPACE, tableName)); + executeViaDriver(String.format("SELECT * from %s.%s WHERE k = 0", KEYSPACE, tableName)); + + // enable it back and do non-partition key queries + enableGuardrail(); + + assertThat(executeSelectViaDriver()).isEmpty(); + } + + private void assertWarnAborts(int warns, int aborts) + { + assertThat(totalWarnings()).as("warnings").isEqualTo(warns); + assertThat(totalAborts()).as("aborts").isEqualTo(aborts); + } + + /** + * Execution of statements via driver will not bypass guardrails as internal queries would do as they are + * done by superuser / they do not have any notion of roles + * + * @return list of warnings + */ + private List<String> executeSelectViaDriver() + { + return executeViaDriver(String.format("SELECT * from %s.%s WHERE v1 = 0", KEYSPACE, tableName)); + } + + private List<String> executeViaDriver(String query) + { + SimpleStatement stmt = new SimpleStatement(query); + stmt.setConsistencyLevel(com.datastax.driver.core.ConsistencyLevel.QUORUM); + ResultSet resultSet = driverSession.execute(stmt); + return resultSet.getExecutionInfo().getWarnings(); + } + + private List<String> executeSelect(long valueToQuery, boolean expectToFail) + { + return cluster.get(1).applyOnInstance((IIsolatedExecutor.SerializableTriFunction<String, String, Long, List<String>>) (keyspace, table, v1) -> { + ClientWarn.instance.captureWarnings(); + CoordinatorWarnings.init(); + + String query = String.format("SELECT * from %s.%s WHERE v1 = %s", keyspace, table, v1); + + try + { + QueryProcessor.execute(query, ALL, QueryState.forInternalCalls()); + if (expectToFail) + fail("expected to fail"); + } + catch (QueryReferencesTooManyIndexesAbortException e) + { + if (!expectToFail) + fail("did not expect to fail"); + + assertTrue(e.nodes >= 1 && e.nodes <= 3); + } + + CoordinatorWarnings.done(); + CoordinatorWarnings.reset(); + return ClientWarn.instance.getWarnings(); + }, KEYSPACE, tableName, valueToQuery); + } + + private void prepareSchema(boolean sai) + { + schemaChange("CREATE TABLE IF NOT EXISTS %s (k bigint, c bigint, v1 bigint, v2 bigint, PRIMARY KEY (k, c));"); + + if (sai) + { + schemaChange("CREATE CUSTOM INDEX IF NOT EXISTS v1_idx ON %s (v1) USING 'StorageAttachedIndex'"); + schemaChange("CREATE CUSTOM INDEX IF NOT EXISTS v2_idx ON %s (v2) USING 'StorageAttachedIndex'"); + } + else + { + schemaChange("CREATE INDEX IF NOT EXISTS v1_idx ON %s (v1)"); + schemaChange("CREATE INDEX IF NOT EXISTS v2_idx ON %s (v2)"); + } + + + for (int i = 1; i < cluster.size() + 1; i++) + { + cluster.get(i).acceptsOnInstance((IIsolatedExecutor.SerializableBiConsumer<String, String>) (ks, tb) -> { + for (ColumnFamilyStore cs : Keyspace.open(ks).getColumnFamilyStores()) + { + if (cs.name.equals(tb)) + { + cs.disableAutoCompaction(); + break; + } + } + }).accept(KEYSPACE, tableName); + } + } + + private void compact(int node) + { + cluster.get(node).acceptsOnInstance((IIsolatedExecutor.SerializableBiConsumer<String, String>) (ks, tb) -> { + for (ColumnFamilyStore cs : Keyspace.open(ks).getColumnFamilyStores()) + { + if (cs.name.equals(tb)) + { + cs.forceMajorCompaction(); + break; + } + } + }).accept(KEYSPACE, tableName); + } + + private long createSSTables(int numberOfSSTables, int... nodesToFlush) + { + Uninterruptibles.sleepUninterruptibly(1, TimeUnit.SECONDS); + + long k = System.currentTimeMillis(); + long value = k + 10; + for (int i = 0; i < numberOfSSTables; i++) + { + // this will be replicated to each node + cluster.coordinator(1) + .execute(format("INSERT INTO %s (k, c, v1, v2) VALUES (?, ?, ?, ?)"), + ConsistencyLevel.ALL, k, k, value, value + 10); + + if (nodesToFlush.length == 0) + cluster.forEach((instance) -> instance.flush(KEYSPACE)); + else + cluster.get(nodesToFlush).forEach((instance) -> instance.flush(KEYSPACE)); + } + + return value; + } + + private void setThresholds(int warn, int fail, int... nodes) + { + Stream<IInvokableInstance> instances; + + if (nodes.length == 0) + instances = cluster.stream(); + else + instances = cluster.get(nodes).stream(); + + instances.forEach(instance -> instance.acceptsOnInstance((IIsolatedExecutor.SerializableBiConsumer<Integer, Integer>) + (w, f) -> Guardrails.instance.setSaiSSTableIndexesPerQueryThreshold(w, f)) + .accept(warn, fail)); + assertTresholds(warn, fail, nodes); + } + + private void enableGuardrail() + { + cluster.forEach(instance -> instance.runOnInstance((IIsolatedExecutor.SerializableRunnable) () -> Guardrails.instance.setNonPartitionRestrictedQueryEnabled(true))); + cluster.forEach(instance -> assertTrue(instance.callsOnInstance((IIsolatedExecutor.SerializableCallable<Boolean>) () -> Guardrails.instance.getNonPartitionRestrictedIndexQueryEnabled()).call())); + } + + private void disableGuardrail() + { + cluster.forEach(instance -> instance.runOnInstance((IIsolatedExecutor.SerializableRunnable) () -> Guardrails.instance.setNonPartitionRestrictedQueryEnabled(false))); + cluster.forEach(instance -> assertFalse(instance.callsOnInstance((IIsolatedExecutor.SerializableCallable<Boolean>) () -> Guardrails.instance.getNonPartitionRestrictedIndexQueryEnabled()).call())); + } + + private void assertTresholds(int expectedWarn, int expectedFail, int... nodes) + { + Stream<IInvokableInstance> instances; + + if (nodes.length == 0) + instances = cluster.stream(); + else + instances = cluster.get(nodes).stream(); + + instances.forEach(instance -> { + assertEquals(expectedWarn, + instance.callsOnInstance((IIsolatedExecutor.SerializableCallable<Integer>) () -> Guardrails.instance.getSaiSSTableIndexesPerQueryWarnThreshold()) + .call().intValue()); + assertEquals(expectedFail, + instance.callsOnInstance((IIsolatedExecutor.SerializableCallable<Integer>) () -> Guardrails.instance.getSaiSSTableIndexesPerQueryFailThreshold()) + .call().intValue()); + }); + } + + private long totalWarnings() + { + return cluster.stream().mapToLong(i -> i.metrics().getCounter("org.apache.cassandra.metrics.keyspace.TooManyIndexReadSSTablesWarnings.distributed_test_keyspace")).sum(); Review Comment: nit: Alternate, more consistent naming ```suggestion return cluster.stream().mapToLong(i -> i.metrics().getCounter("org.apache.cassandra.metrics.keyspace.TooManySSTableIndexesReadWarnings.distributed_test_keyspace")).sum(); ``` -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
