diff --git a/test/distributed/org/apache/cassandra/distributed/upgrade/MixedModeRepairTest.java b/test/distributed/org/apache/cassandra/distributed/upgrade/MixedModeRepairTest.java new file mode 100644 index 0000000000..39f6c9524d --- /dev/null +++ b/test/distributed/org/apache/cassandra/distributed/upgrade/MixedModeRepairTest.java @@ -0,0 +1,126 @@ +/* + * 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.upgrade; + +import java.util.UUID; +import java.util.concurrent.CompletableFuture; +import java.util.concurrent.RejectedExecutionException; +import java.util.concurrent.TimeUnit; +import java.util.concurrent.TimeoutException; + +import org.junit.Assert; +import org.junit.Test; + +import org.apache.cassandra.distributed.UpgradeableCluster; +import org.apache.cassandra.distributed.api.IUpgradeableInstance; + +import static org.apache.cassandra.distributed.api.Feature.GOSSIP; +import static org.apache.cassandra.distributed.api.Feature.NETWORK; +import static org.apache.cassandra.distributed.shared.AssertUtils.assertRows; +import static org.apache.cassandra.distributed.shared.AssertUtils.fail; +import static org.apache.cassandra.distributed.shared.AssertUtils.row; +import static org.apache.cassandra.distributed.shared.Versions.Major; + +public class MixedModeRepairTest extends UpgradeTestBase +{ + public static final int UPGRADED_NODE = 1; + public static final String CREATE_TABLE = withKeyspace("CREATE TABLE %s.t (k uuid, c int, v int, PRIMARY KEY (k, c))"); + public static final String INSERT = withKeyspace("INSERT INTO %s.t (k, c, v) VALUES (?, ?, ?)"); + public static final String SELECT = withKeyspace("SELECT * FROM %s.t WHERE k=?"); + + /** + * Test that repairs fail during a major upgrade. If the repaired node is >= 4.0 thanks to CASSANDRA-13944 there + * will be an informative message. Otherwise, if the repaired node is below 4.0, there won't be such an informative + * message and the repair will take very long to timeout. + */ + @Test + public void testRepairDuringMajorUpgrade() throws Throwable + { + new UpgradeTestBase.TestCase() + .nodes(2) + .nodesToUpgrade(UPGRADED_NODE) + .upgrade(Major.v30, Major.v4) + .upgrade(Major.v3X, Major.v4) + .withConfig(config -> config.with(NETWORK, GOSSIP)) + .setup(cluster -> { + cluster.schemaChange(CREATE_TABLE); + cluster.setUncaughtExceptionsFilter(throwable -> throwable instanceof RejectedExecutionException); + }) + .runAfterNodeUpgrade((cluster, node) -> { + + // run the repair scenario in both the upgraded and the not upgraded node + for (int repairedNode = 1; repairedNode <= cluster.size(); repairedNode++) + { + UUID key = UUID.randomUUID(); + + // only in node 1, create a sstable with a version of a partition + Object[] row1 = row(key, 10, 100); + cluster.get(1).executeInternal(INSERT, row1); + cluster.get(1).flush(KEYSPACE); + + // only in node 2, create a sstable with another version of the partition + Object[] row2 = row(key, 20, 200); + cluster.get(2).executeInternal(INSERT, row2); + cluster.get(2).flush(KEYSPACE); + + // in case of repairing the upgraded node the repair should be rejected with a decriptive error in both + // nodetool output and logs (see CASSANDRA-13944) + if (repairedNode == UPGRADED_NODE) + { + String errorMessage = "Repair is not supported in mixed major version clusters"; + cluster.get(repairedNode) + .nodetoolResult("repair", "--full", KEYSPACE) + .asserts() + .errorContains(errorMessage); + assertLogHas(cluster, repairedNode, errorMessage); + } + // if the node issuing the repair is the not updated node we don't have specific error management, + // so the repair will produce a failure in the upgraded node, and it will take one hour to time out in + // the not upgraded node. Since we don't want to wait that long, we only wait a few seconds for the + // repair before verifying the "unknown verb id" error in the upgraded node. + else + { + try + { + IUpgradeableInstance instance = cluster.get(repairedNode); + CompletableFuture.supplyAsync(() -> instance.nodetoolResult("repair", "--full", KEYSPACE)) + .get(10, TimeUnit.SECONDS); + fail("Repair in the not upgraded node should have timed out"); + } + catch (TimeoutException e) + { + assertLogHas(cluster, UPGRADED_NODE, "unexpected exception caught while processing inbound messages"); + assertLogHas(cluster, UPGRADED_NODE, "java.lang.IllegalArgumentException: Unknown verb id"); + } + } + + // verify that the previous failed repair hasn't repaired the data + assertRows(cluster.get(1).executeInternal(SELECT, key), row1); + assertRows(cluster.get(2).executeInternal(SELECT, key), row2); + } + }) + .run(); + } + + private static void assertLogHas(UpgradeableCluster cluster, int node, String msg) + { + Assert.assertFalse("Unable to find '" + msg + "' in the logs of node " + node, + cluster.get(node).logs().grep(msg).getResult().isEmpty()); + } +}