/**
* 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 com.twitter.distributedlog;
import com.twitter.distributedlog.exceptions.BKTransmitException;
import com.twitter.distributedlog.io.Abortables;
import com.twitter.distributedlog.util.FutureUtils;
import org.apache.bookkeeper.client.BKException;
import org.apache.bookkeeper.proto.BookieServer;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.junit.Test;
import static org.junit.Assert.*;
public class TestFailureAndRecovery extends TestDistributedLogBase {
static final Log LOG = LogFactory.getLog(TestFailureAndRecovery.class);
@Test(timeout = 60000)
public void testSimpleRecovery() throws Exception {
DLMTestUtil.BKLogPartitionWriteHandlerAndClients bkdlmAndClients = createNewBKDLM(conf, "distrlog-simplerecovery");
BKLogSegmentWriter out = bkdlmAndClients.getWriteHandler().startLogSegment(1);
long txid = 1;
for (long i = 1; i <= 100; i++) {
LogRecord op = DLMTestUtil.getLogRecordInstance(txid++);
out.write(op);
if ((i % 10) == 0) {
FutureUtils.result(out.flushAndCommit());
}
}
FutureUtils.result(out.flushAndCommit());
Abortables.abort(out, false);
FutureUtils.result(out.asyncClose());
assertNull(zkc.exists(bkdlmAndClients.getWriteHandler().completedLedgerZNode(1, 100, out.getLogSegmentSequenceNumber()), false));
assertNotNull(zkc.exists(bkdlmAndClients.getWriteHandler().inprogressZNode(out.getLogSegmentId(), 1, out.getLogSegmentSequenceNumber()), false));
FutureUtils.result(bkdlmAndClients.getWriteHandler().recoverIncompleteLogSegments());
assertNotNull(zkc.exists(bkdlmAndClients.getWriteHandler().completedLedgerZNode(1, 100, out.getLogSegmentSequenceNumber()), false));
assertNull(zkc.exists(bkdlmAndClients.getWriteHandler().inprogressZNode(out.getLogSegmentId(), 1, out.getLogSegmentSequenceNumber()), false));
}
/**
* Test that if enough bookies fail to prevent an ensemble,
* writes the bookkeeper will fail. Test that when once again
* an ensemble is available, it can continue to write.
*/
@Test(timeout = 60000)
public void testAllBookieFailure() throws Exception {
BookieServer bookieToFail = bkutil.newBookie();
BookieServer replacementBookie = null;
try {
int ensembleSize = numBookies + 1;
assertEquals("Begin: New bookie didn't start",
ensembleSize, bkutil.checkBookiesUp(ensembleSize, 10));
// ensure that the journal manager has to use all bookies,
// so that a failure will fail the journal manager
DistributedLogConfiguration conf = new DistributedLogConfiguration();
conf.setEnsembleSize(ensembleSize);
conf.setWriteQuorumSize(ensembleSize);
conf.setAckQuorumSize(ensembleSize);
long txid = 1;
DLMTestUtil.BKLogPartitionWriteHandlerAndClients bkdlmAndClients = createNewBKDLM(conf, "distrlog-allbookiefailure");
BKLogSegmentWriter out = bkdlmAndClients.getWriteHandler().startLogSegment(txid);
for (long i = 1; i <= 3; i++) {
LogRecord op = DLMTestUtil.getLogRecordInstance(txid++);
out.write(op);
}
FutureUtils.result(out.flushAndCommit());
bookieToFail.shutdown();
assertEquals("New bookie didn't die",
numBookies, bkutil.checkBookiesUp(numBookies, 10));
try {
for (long i = 1; i <= 3; i++) {
LogRecord op = DLMTestUtil.getLogRecordInstance(txid++);
out.write(op);
txid++;
}
FutureUtils.result(out.flushAndCommit());
fail("should not get to this stage");
} catch (BKTransmitException bkte) {
LOG.debug("Error writing to bookkeeper", bkte);
assertEquals("Invalid exception message",
BKException.Code.NotEnoughBookiesException, bkte.getBKResultCode());
}
replacementBookie = bkutil.newBookie();
assertEquals("Replacement: New bookie didn't start",
numBookies + 1, bkutil.checkBookiesUp(numBookies + 1, 10));
out = bkdlmAndClients.getWriteHandler().startLogSegment(txid);
for (long i = 1; i <= 3; i++) {
LogRecord op = DLMTestUtil.getLogRecordInstance(txid++);
out.write(op);
}
FutureUtils.result(out.flushAndCommit());
} catch (Exception e) {
LOG.error("Exception in test", e);
throw e;
} finally {
if (replacementBookie != null) {
replacementBookie.shutdown();
}
bookieToFail.shutdown();
if (bkutil.checkBookiesUp(numBookies, 30) != numBookies) {
LOG.warn("Not all bookies from this test shut down, expect errors");
}
}
}
/**
* Test that a BookKeeper JM can continue to work across the
* failure of a bookie. This should be handled transparently
* by bookkeeper.
*/
@Test(timeout = 60000)
public void testOneBookieFailure() throws Exception {
BookieServer bookieToFail = bkutil.newBookie();
BookieServer replacementBookie = null;
try {
int ensembleSize = numBookies + 1;
assertEquals("New bookie didn't start",
ensembleSize, bkutil.checkBookiesUp(ensembleSize, 10));
// ensure that the journal manager has to use all bookies,
// so that a failure will fail the journal manager
DistributedLogConfiguration conf = new DistributedLogConfiguration();
conf.setEnsembleSize(ensembleSize);
conf.setWriteQuorumSize(ensembleSize);
conf.setAckQuorumSize(ensembleSize);
long txid = 1;
DLMTestUtil.BKLogPartitionWriteHandlerAndClients bkdlmAndClients = createNewBKDLM(conf, "distrlog-onebookiefailure");
BKLogSegmentWriter out = bkdlmAndClients.getWriteHandler().startLogSegment(txid);
for (long i = 1; i <= 3; i++) {
LogRecord op = DLMTestUtil.getLogRecordInstance(txid++);
out.write(op);
}
FutureUtils.result(out.flushAndCommit());
replacementBookie = bkutil.newBookie();
assertEquals("replacement bookie didn't start",
ensembleSize + 1, bkutil.checkBookiesUp(ensembleSize + 1, 10));
bookieToFail.shutdown();
assertEquals("New bookie didn't die",
ensembleSize, bkutil.checkBookiesUp(ensembleSize, 10));
for (long i = 1; i <= 3; i++) {
LogRecord op = DLMTestUtil.getLogRecordInstance(txid++);
out.write(op);
}
FutureUtils.result(out.flushAndCommit());
} catch (Exception e) {
LOG.error("Exception in test", e);
throw e;
} finally {
if (replacementBookie != null) {
replacementBookie.shutdown();
}
bookieToFail.shutdown();
if (bkutil.checkBookiesUp(numBookies, 30) != numBookies) {
LOG.warn("Not all bookies from this test shut down, expect errors");
}
}
}
@Test(timeout = 60000)
public void testRecoveryEmptyLedger() throws Exception {
DLMTestUtil.BKLogPartitionWriteHandlerAndClients bkdlmAndClients = createNewBKDLM(conf, "distrlog-recovery-empty-ledger");
BKLogSegmentWriter out = bkdlmAndClients.getWriteHandler().startLogSegment(1);
long txid = 1;
for (long i = 1; i <= 100; i++) {
LogRecord op = DLMTestUtil.getLogRecordInstance(txid++);
out.write(op);
if ((i % 10) == 0) {
FutureUtils.result(out.flushAndCommit());
}
}
FutureUtils.result(out.flushAndCommit());
FutureUtils.result(out.asyncClose());
bkdlmAndClients.getWriteHandler().completeAndCloseLogSegment(out.getLogSegmentSequenceNumber(), out.getLogSegmentId(), 1, 100, 100);
assertNotNull(zkc.exists(bkdlmAndClients.getWriteHandler().completedLedgerZNode(1, 100, out.getLogSegmentSequenceNumber()), false));
BKLogSegmentWriter outEmpty = bkdlmAndClients.getWriteHandler().startLogSegment(101);
Abortables.abort(outEmpty, false);
assertNull(zkc.exists(bkdlmAndClients.getWriteHandler().completedLedgerZNode(101, 101, outEmpty.getLogSegmentSequenceNumber()), false));
assertNotNull(zkc.exists(bkdlmAndClients.getWriteHandler().inprogressZNode(outEmpty.getLogSegmentId(), 101, outEmpty.getLogSegmentSequenceNumber()), false));
FutureUtils.result(bkdlmAndClients.getWriteHandler().recoverIncompleteLogSegments());
assertNull(zkc.exists(bkdlmAndClients.getWriteHandler().inprogressZNode(outEmpty.getLogSegmentId(), outEmpty.getLogSegmentSequenceNumber(), 101), false));
assertNotNull(zkc.exists(bkdlmAndClients.getWriteHandler().completedLedgerZNode(101, 101, outEmpty.getLogSegmentSequenceNumber()), false));
}
@Test(timeout = 60000)
public void testRecoveryAPI() throws Exception {
DistributedLogManager dlm = createNewDLM(conf, "distrlog-recovery-api");
BKSyncLogWriter out = (BKSyncLogWriter) dlm.startLogSegmentNonPartitioned();
long txid = 1;
for (long i = 1; i <= 100; i++) {
LogRecord op = DLMTestUtil.getLogRecordInstance(txid++);
out.write(op);
if ((i % 10) == 0) {
out.setReadyToFlush();
out.flushAndSync();
}
}
BKLogSegmentWriter perStreamLogWriter = out.getCachedLogWriter();
out.setReadyToFlush();
out.flushAndSync();
out.abort();
BKLogWriteHandler blplm1 = ((BKDistributedLogManager) (dlm)).createWriteHandler(true);
assertNull(zkc.exists(blplm1.completedLedgerZNode(1, 100,
perStreamLogWriter.getLogSegmentSequenceNumber()), false));
assertNotNull(zkc.exists(blplm1.inprogressZNode(perStreamLogWriter.getLogSegmentId(), 1,
perStreamLogWriter.getLogSegmentSequenceNumber()), false));
dlm.recover();
assertNotNull(zkc.exists(blplm1.completedLedgerZNode(1, 100,
perStreamLogWriter.getLogSegmentSequenceNumber()), false));
assertNull(zkc.exists(blplm1.inprogressZNode(perStreamLogWriter.getLogSegmentId(), 1,
perStreamLogWriter.getLogSegmentSequenceNumber()), false));
FutureUtils.result(blplm1.asyncClose());
assertEquals(100, dlm.getLogRecordCount());
dlm.close();
}
}