Skip to content

fix(spanner): trigger best-effort rollback on indeterminate commit failures #14831

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
sakthivelmanii wants to merge 2 commits into
base: main
Choose a base branch
from
Open

fix(spanner): trigger best-effort rollback on indeterminate commit failures #14831

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
3aa7674
fix(spanner): trigger best-effort rollback on indeterminate commit fa…
sakthivelmanii Oct 29, 2026
4553a6b
chore: format TransactionManagerImplTest with google-java-format
sakthivelmanii Oct 29, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions ...gle-cloud-spanner/src/main/java/com/google/cloud/spanner/AsyncTransactionManagerImpl.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -165,6 +165,9 @@ public void onFailure(Throwable t) {
txnState = TransactionState.ABORTED;
} else {
txnState = TransactionState.COMMIT_FAILED;
if (TransactionRunnerImpl.shouldRollbackOnCommitFailure(t)) {
TransactionRunnerImpl.rollbackBestEffort(txn, t);
}
if (span != null) {
span.setStatus(t);
span.end();
Expand Down
3 changes: 3 additions & 0 deletions ...r/google-cloud-spanner/src/main/java/com/google/cloud/spanner/TransactionManagerImpl.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -92,6 +92,9 @@ public void commit() {
throw e1;
} catch (SpannerException e2) {
txnState = TransactionState.COMMIT_FAILED;
if (TransactionRunnerImpl.shouldRollbackOnCommitFailure(e2)) {
TransactionRunnerImpl.rollbackBestEffort(txn, e2);
}
throw e2;
} finally {
// At this point, if the TransactionState is not ABORTED, then the transaction has reached an
Expand Down
25 changes: 25 additions & 0 deletions ...er/google-cloud-spanner/src/main/java/com/google/cloud/spanner/TransactionRunnerImpl.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1404,6 +1404,9 @@ private <T> T runInternal(final TransactionCallable<T> txCallable) {
"Transaction Attempt Failed in Commit",
ImmutableMap.of(
"Attempt", attempt.longValue(), "Status", e.getErrorCode().toString()));
if (shouldRollbackOnCommitFailure(e)) {
rollbackBestEffort(txn, e);
}
throw e;
}
};
Expand All @@ -1428,4 +1431,26 @@ public void invalidate() {

@Override
public void close() {}

static boolean shouldRollbackOnCommitFailure(Throwable t) {
SpannerException e = SpannerExceptionFactory.asSpannerException(t);
switch (e.getErrorCode()) {
case DEADLINE_EXCEEDED:
case CANCELLED:
case UNAVAILABLE:
return true;
case RESOURCE_EXHAUSTED:
return e.getRetryDelayInMillis() > -1L;
default:
return false;
}
}

static void rollbackBestEffort(TransactionContextImpl txn, Throwable commitError) {
try {
ApiFuture<Empty> unused = txn.rollbackAsync();
} catch (RuntimeException rollbackError) {
commitError.addSuppressed(rollbackError);
}
}
}
118 changes: 118 additions & 0 deletions ...cloud-spanner/src/test/java/com/google/cloud/spanner/AsyncTransactionManagerImplTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,16 +16,23 @@

package com.google.cloud.spanner;

import static org.junit.Assert.assertEquals;
import static org.junit.Assert.assertThrows;
import static org.mockito.ArgumentMatchers.any;
import static org.mockito.ArgumentMatchers.eq;
import static org.mockito.Mockito.mock;
import static org.mockito.Mockito.never;
import static org.mockito.Mockito.times;
import static org.mockito.Mockito.verify;
import static org.mockito.Mockito.when;

import com.google.api.core.ApiFutures;
import com.google.cloud.Timestamp;
import com.google.cloud.spanner.TransactionManager.TransactionState;
import com.google.protobuf.Empty;
import io.opentelemetry.api.trace.Span;
import io.opentelemetry.context.Scope;
import java.util.concurrent.ExecutionException;
import org.junit.Test;
import org.junit.runner.RunWith;
import org.mockito.Mock;
Expand Down Expand Up @@ -56,4 +63,115 @@ public void testCommitReturnsCommitStats() {
verify(transaction).commitAsync();
}
}

@Test
public void testCommitFailsWithDeadlineExceededRollsBack() {
Span oTspan = mock(Span.class);
ISpan span = new OpenTelemetrySpan(oTspan);
when(oTspan.makeCurrent()).thenReturn(mock(Scope.class));
when(session.newTransaction(eq(Options.fromTransactionOptions()), any()))
.thenReturn(transaction);
when(transaction.ensureTxnAsync()).thenReturn(ApiFutures.immediateFuture(null));
when(transaction.commitAsync())
.thenReturn(
ApiFutures.immediateFailedFuture(
SpannerExceptionFactory.newSpannerException(
ErrorCode.DEADLINE_EXCEEDED, "commit deadline exceeded")));
when(transaction.rollbackAsync())
.thenReturn(ApiFutures.immediateFuture(Empty.getDefaultInstance()));
try (AsyncTransactionManagerImpl manager = new AsyncTransactionManagerImpl(session, span)) {
manager.beginAsync();
assertThrows(ExecutionException.class, () -> manager.commitAsync().get());
assertEquals(TransactionState.COMMIT_FAILED, manager.getState());
verify(transaction, times(1)).rollbackAsync();
}
}

@Test
public void testCommitFailsWithCancelledRollsBack() {
Span oTspan = mock(Span.class);
ISpan span = new OpenTelemetrySpan(oTspan);
when(oTspan.makeCurrent()).thenReturn(mock(Scope.class));
when(session.newTransaction(eq(Options.fromTransactionOptions()), any()))
.thenReturn(transaction);
when(transaction.ensureTxnAsync()).thenReturn(ApiFutures.immediateFuture(null));
when(transaction.commitAsync())
.thenReturn(
ApiFutures.immediateFailedFuture(
SpannerExceptionFactory.newSpannerException(
ErrorCode.CANCELLED, "commit cancelled")));
when(transaction.rollbackAsync())
.thenReturn(ApiFutures.immediateFuture(Empty.getDefaultInstance()));
try (AsyncTransactionManagerImpl manager = new AsyncTransactionManagerImpl(session, span)) {
manager.beginAsync();
assertThrows(ExecutionException.class, () -> manager.commitAsync().get());
assertEquals(TransactionState.COMMIT_FAILED, manager.getState());
verify(transaction, times(1)).rollbackAsync();
}
}

@Test
public void testCommitFailsWithUnavailableRollsBack() {
Span oTspan = mock(Span.class);
ISpan span = new OpenTelemetrySpan(oTspan);
when(oTspan.makeCurrent()).thenReturn(mock(Scope.class));
when(session.newTransaction(eq(Options.fromTransactionOptions()), any()))
.thenReturn(transaction);
when(transaction.ensureTxnAsync()).thenReturn(ApiFutures.immediateFuture(null));
when(transaction.commitAsync())
.thenReturn(
ApiFutures.immediateFailedFuture(
SpannerExceptionFactory.newSpannerException(
ErrorCode.UNAVAILABLE, "commit unavailable")));
when(transaction.rollbackAsync())
.thenReturn(ApiFutures.immediateFuture(Empty.getDefaultInstance()));
try (AsyncTransactionManagerImpl manager = new AsyncTransactionManagerImpl(session, span)) {
manager.beginAsync();
assertThrows(ExecutionException.class, () -> manager.commitAsync().get());
assertEquals(TransactionState.COMMIT_FAILED, manager.getState());
verify(transaction, times(1)).rollbackAsync();
}
}

@Test
public void testCommitFailsWithResourceExhaustedWithRetryInfoRollsBack() {
Span oTspan = mock(Span.class);
ISpan span = new OpenTelemetrySpan(oTspan);
when(oTspan.makeCurrent()).thenReturn(mock(Scope.class));
when(session.newTransaction(eq(Options.fromTransactionOptions()), any()))
.thenReturn(transaction);
when(transaction.ensureTxnAsync()).thenReturn(ApiFutures.immediateFuture(null));
when(transaction.commitAsync())
.thenReturn(
ApiFutures.immediateFailedFuture(
TransactionRunnerImplTest.resourceExhaustedWithRetryInfo()));
when(transaction.rollbackAsync())
.thenReturn(ApiFutures.immediateFuture(Empty.getDefaultInstance()));
try (AsyncTransactionManagerImpl manager = new AsyncTransactionManagerImpl(session, span)) {
manager.beginAsync();
assertThrows(ExecutionException.class, () -> manager.commitAsync().get());
assertEquals(TransactionState.COMMIT_FAILED, manager.getState());
verify(transaction, times(1)).rollbackAsync();
}
}

@Test
public void testCommitFailsWithNonAbortDoesNotRollBack() {
Span oTspan = mock(Span.class);
ISpan span = new OpenTelemetrySpan(oTspan);
when(oTspan.makeCurrent()).thenReturn(mock(Scope.class));
when(session.newTransaction(eq(Options.fromTransactionOptions()), any()))
.thenReturn(transaction);
when(transaction.ensureTxnAsync()).thenReturn(ApiFutures.immediateFuture(null));
when(transaction.commitAsync())
.thenReturn(
ApiFutures.immediateFailedFuture(
SpannerExceptionFactory.newSpannerException(ErrorCode.UNKNOWN, "unknown error")));
try (AsyncTransactionManagerImpl manager = new AsyncTransactionManagerImpl(session, span)) {
manager.beginAsync();
assertThrows(ExecutionException.class, () -> manager.commitAsync().get());
assertEquals(TransactionState.COMMIT_FAILED, manager.getState());
}
verify(transaction, never()).rollbackAsync();
}
}
112 changes: 112 additions & 0 deletions .../google-cloud-spanner/src/test/java/com/google/cloud/spanner/OrphanedTransactionTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,21 +16,26 @@

package com.google.cloud.spanner;

import static org.junit.Assert.assertEquals;
import static org.junit.Assert.assertNull;
import static org.junit.Assert.assertThrows;

import com.google.api.core.ApiFuture;
import com.google.cloud.NoCredentials;
import com.google.cloud.spanner.AsyncTransactionManager.CommitTimestampFuture;
import com.google.cloud.spanner.AsyncTransactionManager.TransactionContextFuture;
import com.google.cloud.spanner.MockSpannerServiceImpl.SimulatedExecutionTime;
import com.google.cloud.spanner.MockSpannerServiceImpl.StatementResult;
import com.google.cloud.spanner.connection.AbstractMockServerTest;
import com.google.cloud.spanner.connection.RandomResultSetGenerator;
import com.google.common.base.Function;
import com.google.common.util.concurrent.MoreExecutors;
import com.google.spanner.v1.ExecuteSqlRequest;
import com.google.spanner.v1.RollbackRequest;
import io.grpc.ManagedChannelBuilder;
import io.grpc.Status;
import java.util.Objects;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import org.junit.BeforeClass;
Expand Down Expand Up @@ -144,4 +149,111 @@ public void testOrphanedTransactionWithFailedFirstQuery() throws Exception {
executor.shutdown();
}
}

private void assertRollbackRequestReceived() throws InterruptedException {
long deadline = System.currentTimeMillis() + 200L;
while (mockSpanner.countRequestsOfType(RollbackRequest.class) == 0
&& System.currentTimeMillis() < deadline) {
Thread.sleep(10L);
}
assertEquals(
"Expected RollbackRequest after Commit DEADLINE_EXCEEDED",
1,
mockSpanner.countRequestsOfType(RollbackRequest.class));
}

@Test
public void testTransactionRunnerCommitDeadlineExceededSendsRollback() throws Exception {
mockSpanner.clearRequests();
mockSpanner.setCommitExecutionTime(
SimulatedExecutionTime.ofException(
Status.DEADLINE_EXCEEDED
.withDescription("commit deadline exceeded")
.asRuntimeException()));
try (Spanner spanner = createSpanner()) {
DatabaseClient client =
spanner.getDatabaseClient(
DatabaseId.of("fake-project", "fake-instance", "fake-database"));
SpannerException e =
assertThrows(
SpannerException.class,
() ->
client
.readWriteTransaction()
.run(
txn -> {
try (ResultSet rs = txn.executeQuery(STATEMENT)) {
while (rs.next()) {}
}
return null;
}));
assertEquals(ErrorCode.DEADLINE_EXCEEDED, e.getErrorCode());
assertRollbackRequestReceived();
} finally {
mockSpanner.removeAllExecutionTimes();
}
}

@Test
public void testTransactionManagerCommitDeadlineExceededSendsRollback() throws Exception {
mockSpanner.clearRequests();
mockSpanner.setCommitExecutionTime(
SimulatedExecutionTime.ofException(
Status.DEADLINE_EXCEEDED
.withDescription("commit deadline exceeded")
.asRuntimeException()));
try (Spanner spanner = createSpanner()) {
DatabaseClient client =
spanner.getDatabaseClient(
DatabaseId.of("fake-project", "fake-instance", "fake-database"));
try (TransactionManager manager = client.transactionManager()) {
TransactionContext txn = manager.begin();
try (ResultSet rs = txn.executeQuery(STATEMENT)) {
while (rs.next()) {}
}
SpannerException e = assertThrows(SpannerException.class, manager::commit);
assertEquals(ErrorCode.DEADLINE_EXCEEDED, e.getErrorCode());
}
assertRollbackRequestReceived();
} finally {
mockSpanner.removeAllExecutionTimes();
}
}

@Test
public void testAsyncTransactionManagerCommitDeadlineExceededSendsRollback() throws Exception {
mockSpanner.clearRequests();
mockSpanner.setCommitExecutionTime(
SimulatedExecutionTime.ofException(
Status.DEADLINE_EXCEEDED
.withDescription("commit deadline exceeded")
.asRuntimeException()));
try (Spanner spanner = createSpanner()) {
DatabaseClient client =
spanner.getDatabaseClient(
DatabaseId.of("fake-project", "fake-instance", "fake-database"));
try (AsyncTransactionManager manager = client.transactionManagerAsync()) {
TransactionContextFuture context = manager.beginAsync();
CommitTimestampFuture commitFuture =
context
.then(
(txn, input) -> {
try (AsyncResultSet rs = txn.executeQueryAsync(STATEMENT)) {
return rs.toListAsync(
row -> Objects.requireNonNull(row).getValue(0).getAsString(),
MoreExecutors.directExecutor());
}
},
MoreExecutors.directExecutor())
.commitAsync();
ExecutionException e = assertThrows(ExecutionException.class, commitFuture::get);
assertEquals(
ErrorCode.DEADLINE_EXCEEDED,
SpannerExceptionFactory.asSpannerException(e.getCause()).getErrorCode());
}
assertRollbackRequestReceived();
} finally {
mockSpanner.removeAllExecutionTimes();
}
}
}
Loading
Loading