sashapolo commented on code in PR #3447:
URL: https://github.com/apache/ignite-3/pull/3447#discussion_r1533694913
##########
modules/storage-api/src/testFixtures/java/org/apache/ignite/internal/storage/index/impl/AbstractTestIndexStorage.java:
##########
@@ -55,7 +56,7 @@ public int pendingCursors() {
@Override
public Cursor<RowId> get(BinaryTuple key) {
- checkStorageClosedOrInProcessOfRebalance();
+ checkStorageClosedOrInProcessOfRebalance(true);
Review Comment:
Shall we create a wrapper over index storages (e.g.
`RebalanceAwareIndexStorage`), that will perform all these checks? It's a bit
frustrating that every storage implementation has to check this
##########
modules/table/src/main/java/org/apache/ignite/internal/table/distributed/replicator/PartitionReplicaListener.java:
##########
@@ -392,6 +395,23 @@ public PartitionReplicaListener(
prepareIndexBuilderTxRwOperationTracker();
}
+ private static void rethrowTranslated(Throwable ex) {
+ Throwable unwrapped = ex;
Review Comment:
Can we use `ExceptionUtils.unwrapCause`?
##########
modules/index/src/integrationTest/java/org/apache/ignite/internal/index/ItIndexAndIndexStorageDestructionTest.java:
##########
@@ -0,0 +1,157 @@
+/*
+ * 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.ignite.internal.index;
+
+import static org.apache.ignite.lang.ErrorGroups.Sql.STALE_PLAN_ERR;
+import static org.hamcrest.MatcherAssert.assertThat;
+import static org.hamcrest.Matchers.is;
+import static org.hamcrest.Matchers.notNullValue;
+import static org.junit.jupiter.api.Assertions.assertDoesNotThrow;
+import static org.junit.jupiter.api.Assertions.assertThrows;
+import static org.junit.jupiter.api.Assertions.fail;
+
+import org.apache.ignite.internal.ClusterPerTestIntegrationTest;
+import org.apache.ignite.internal.app.IgniteImpl;
+import org.apache.ignite.internal.catalog.descriptors.CatalogIndexDescriptor;
+import org.apache.ignite.internal.storage.index.IndexStorage;
+import
org.apache.ignite.internal.storage.pagememory.index.AbstractPageMemoryIndexStorage;
+import
org.apache.ignite.internal.storage.rocksdb.index.AbstractRocksDbIndexStorage;
+import org.apache.ignite.internal.table.TableImpl;
+import org.apache.ignite.internal.table.distributed.TableIndexStoragesSupplier;
+import
org.apache.ignite.internal.table.distributed.TableSchemaAwareIndexStorage;
+import org.apache.ignite.lang.Cursor;
+import org.apache.ignite.sql.SqlException;
+import org.apache.ignite.table.KeyValueView;
+import org.apache.ignite.table.Tuple;
+import org.apache.ignite.table.criteria.CriteriaException;
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.Test;
+
+/**
+ * Tests about accessing destroyed index storages.
+ */
+class ItIndexAndIndexStorageDestructionTest extends
ClusterPerTestIntegrationTest {
+ private static final String TABLE_NAME = "TEST_TABLE";
+ private static final String INDEX_NAME = "TEST_INDEX";
+
+ private static final int PREEXISTING_KEY = 1;
+ private static final int ANOTHER_KEY = 2;
+
+ private IgniteImpl node;
+
+ @Override
+ protected int initialNodes() {
+ return 1;
+ }
+
+ @BeforeEach
+ void createTableAndIndex() {
+ node = cluster.aliveNode();
+
+ cluster.doInSession(0, session -> {
+ session.execute(null, "CREATE TABLE " + TABLE_NAME + " (id INT
PRIMARY KEY, name VARCHAR)");
+ session.execute(null, "CREATE INDEX " + INDEX_NAME + " ON " +
TABLE_NAME + "(name)");
+
+ session.execute(null, "INSERT INTO " + TABLE_NAME + " (id, name)
VALUES (" + PREEXISTING_KEY + ", 'John')");
+ });
+
+ TableImpl table = (TableImpl) node.tables().table(TABLE_NAME);
+
+ initiateIndexStoragesDestruction(table, INDEX_NAME);
+ }
+
+ private void initiateIndexStoragesDestruction(TableImpl table, String
indexName) {
+ int indexId = indexId(indexName);
+
+ for (int partitionNumber = 0; partitionNumber <
table.internalTable().partitions(); partitionNumber++) {
+ TableIndexStoragesSupplier storagesSupplier =
table.indexStorageAdapters(partitionNumber);
+ TableSchemaAwareIndexStorage schemaAwareStorage =
storagesSupplier.get().get(indexId);
+ IndexStorage indexStorage = schemaAwareStorage.storage();
+
+ if (indexStorage instanceof AbstractPageMemoryIndexStorage) {
+ ((AbstractPageMemoryIndexStorage<?, ?>)
indexStorage).transitionToDestroyedState();
+ } else if (indexStorage instanceof AbstractRocksDbIndexStorage) {
+ ((AbstractRocksDbIndexStorage)
indexStorage).transitionToDestroyedState();
+ } else {
+ fail("Index storage is not of a type supported by the test: "
+ indexStorage.getClass());
+ }
+ }
+ }
+
+ private int indexId(String indexName) {
+ CatalogIndexDescriptor indexDescriptor =
node.catalogManager().aliveIndex(indexName, node.clock().nowLong());
+ assertThat(indexDescriptor, is(notNullValue()));
+
+ return indexDescriptor.id();
+ }
+
+ @Test
+ void insertTouchingDestroyedIndexStorageDoesNotFailWholeWrite() {
+ KeyValueView<Integer, String> view = node.tables().table(TABLE_NAME)
Review Comment:
How do you check that this call touches a destroyed index?
##########
modules/table/src/main/java/org/apache/ignite/internal/table/distributed/replicator/PartitionReplicaListener.java:
##########
@@ -402,6 +422,12 @@ public CompletableFuture<ReplicaResult>
invoke(ReplicaRequest request, String se
} else {
return new ReplicaResult(res, null);
}
+ })
+ .exceptionally(ex -> {
+ rethrowTranslated(ex);
Review Comment:
A neat approach is to actually make `rethrowTranslated` return an exception
and throw it here instead. This way you won't have to write `return null` down
below, for example `throw translateException(ex);`
##########
modules/storage-page-memory/src/main/java/org/apache/ignite/internal/storage/pagememory/index/AbstractPageMemoryIndexStorage.java:
##########
@@ -108,7 +109,7 @@ protected AbstractPageMemoryIndexStorage(
@Override
public @Nullable RowId getNextRowIdToBuild() {
- return busy(() -> {
+ return busyNonRead(() -> {
Review Comment:
Why is this a `nonRead`?
##########
modules/core/src/main/java/org/apache/ignite/internal/table/criteria/CriteriaExceptionMapperUtil.java:
##########
@@ -57,10 +57,6 @@ public static Throwable
mapToPublicCriteriaException(Throwable origin) {
return e;
}
- if (e instanceof SqlException) {
- return new CriteriaException(INTERNAL_ERR, e);
Review Comment:
Why is this change needed? Does this exception hide something?
##########
modules/index/src/integrationTest/java/org/apache/ignite/internal/index/ItIndexAndIndexStorageDestructionTest.java:
##########
@@ -0,0 +1,157 @@
+/*
+ * 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.ignite.internal.index;
+
+import static org.apache.ignite.lang.ErrorGroups.Sql.STALE_PLAN_ERR;
+import static org.hamcrest.MatcherAssert.assertThat;
+import static org.hamcrest.Matchers.is;
+import static org.hamcrest.Matchers.notNullValue;
+import static org.junit.jupiter.api.Assertions.assertDoesNotThrow;
+import static org.junit.jupiter.api.Assertions.assertThrows;
+import static org.junit.jupiter.api.Assertions.fail;
+
+import org.apache.ignite.internal.ClusterPerTestIntegrationTest;
+import org.apache.ignite.internal.app.IgniteImpl;
+import org.apache.ignite.internal.catalog.descriptors.CatalogIndexDescriptor;
+import org.apache.ignite.internal.storage.index.IndexStorage;
+import
org.apache.ignite.internal.storage.pagememory.index.AbstractPageMemoryIndexStorage;
+import
org.apache.ignite.internal.storage.rocksdb.index.AbstractRocksDbIndexStorage;
+import org.apache.ignite.internal.table.TableImpl;
+import org.apache.ignite.internal.table.distributed.TableIndexStoragesSupplier;
+import
org.apache.ignite.internal.table.distributed.TableSchemaAwareIndexStorage;
+import org.apache.ignite.lang.Cursor;
+import org.apache.ignite.sql.SqlException;
+import org.apache.ignite.table.KeyValueView;
+import org.apache.ignite.table.Tuple;
+import org.apache.ignite.table.criteria.CriteriaException;
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.Test;
+
+/**
+ * Tests about accessing destroyed index storages.
+ */
+class ItIndexAndIndexStorageDestructionTest extends
ClusterPerTestIntegrationTest {
+ private static final String TABLE_NAME = "TEST_TABLE";
+ private static final String INDEX_NAME = "TEST_INDEX";
+
+ private static final int PREEXISTING_KEY = 1;
+ private static final int ANOTHER_KEY = 2;
+
+ private IgniteImpl node;
+
+ @Override
+ protected int initialNodes() {
+ return 1;
+ }
+
+ @BeforeEach
+ void createTableAndIndex() {
+ node = cluster.aliveNode();
+
+ cluster.doInSession(0, session -> {
+ session.execute(null, "CREATE TABLE " + TABLE_NAME + " (id INT
PRIMARY KEY, name VARCHAR)");
+ session.execute(null, "CREATE INDEX " + INDEX_NAME + " ON " +
TABLE_NAME + "(name)");
+
+ session.execute(null, "INSERT INTO " + TABLE_NAME + " (id, name)
VALUES (" + PREEXISTING_KEY + ", 'John')");
+ });
+
+ TableImpl table = (TableImpl) node.tables().table(TABLE_NAME);
+
+ initiateIndexStoragesDestruction(table, INDEX_NAME);
+ }
+
+ private void initiateIndexStoragesDestruction(TableImpl table, String
indexName) {
+ int indexId = indexId(indexName);
+
+ for (int partitionNumber = 0; partitionNumber <
table.internalTable().partitions(); partitionNumber++) {
+ TableIndexStoragesSupplier storagesSupplier =
table.indexStorageAdapters(partitionNumber);
+ TableSchemaAwareIndexStorage schemaAwareStorage =
storagesSupplier.get().get(indexId);
+ IndexStorage indexStorage = schemaAwareStorage.storage();
+
+ if (indexStorage instanceof AbstractPageMemoryIndexStorage) {
+ ((AbstractPageMemoryIndexStorage<?, ?>)
indexStorage).transitionToDestroyedState();
+ } else if (indexStorage instanceof AbstractRocksDbIndexStorage) {
+ ((AbstractRocksDbIndexStorage)
indexStorage).transitionToDestroyedState();
+ } else {
+ fail("Index storage is not of a type supported by the test: "
+ indexStorage.getClass());
+ }
+ }
+ }
+
+ private int indexId(String indexName) {
+ CatalogIndexDescriptor indexDescriptor =
node.catalogManager().aliveIndex(indexName, node.clock().nowLong());
+ assertThat(indexDescriptor, is(notNullValue()));
+
+ return indexDescriptor.id();
+ }
+
+ @Test
+ void insertTouchingDestroyedIndexStorageDoesNotFailWholeWrite() {
+ KeyValueView<Integer, String> view = node.tables().table(TABLE_NAME)
+ .keyValueView(Integer.class, String.class);
+
+ assertDoesNotThrow(() -> view.put(null, ANOTHER_KEY, "Mary"));
+
+ assertThat(view.get(null, ANOTHER_KEY), is("Mary"));
+ }
+
+ @Test
+ void updateTouchingDestroyedIndexStorageDoesNotFailWholeWrite() {
+ KeyValueView<Integer, String> view = node.tables().table(TABLE_NAME)
+ .keyValueView(Integer.class, String.class);
+
+ assertDoesNotThrow(() -> view.put(null, PREEXISTING_KEY, "Mary"));
+
+ assertThat(view.get(null, PREEXISTING_KEY), is("Mary"));
+ }
+
+ @Test
+ void replaceTouchingDestroyedIndexStorageDoesNotFailWholeWrite() {
+ KeyValueView<Integer, String> view = node.tables().table(TABLE_NAME)
+ .keyValueView(Integer.class, String.class);
+
+ assertDoesNotThrow(() -> view.replace(null, PREEXISTING_KEY, "John",
"Mary"));
+
+ assertThat(view.get(null, PREEXISTING_KEY), is("Mary"));
+ }
+
+ @Test
+ void sqlReadFromDestroyedIndexStorageFailsWithStalePlanError() {
+ SqlException ex = assertThrows(
+ SqlException.class,
+ () -> cluster.query(0, "SELECT * FROM " + TABLE_NAME + " WHERE
name = 'John'", rs -> null)
+ );
+
+ assertThat(ex.code(), is(STALE_PLAN_ERR));
+ assertThat(ex.getMessage(), is("The plan is stale. Please retry."));
+ }
+
+ @Test
+ void kvQueryFromDestroyedIndexStorageFailsWithStalePlanError() {
+ KeyValueView<Tuple, Tuple> view =
node.tables().table(TABLE_NAME).keyValueView();
+
+ CriteriaException ex = assertThrows(CriteriaException.class, () -> {
+ try (Cursor<?> cursor = view.query(null, null, INDEX_NAME)) {
+ cursor.next();
+ }
+ });
+
+ assertThat(ex.code(), is(STALE_PLAN_ERR));
+ assertThat(ex.getMessage(), is("The plan is stale. Please retry."));
Review Comment:
This error message looks strange. Who is it intended for?
--
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]
