HDFS-9329. TestBootstrapStandby#testRateThrottling is flaky because fsimage size is smaller than IO buffer size. Contributed by Zhe Zhang.

Conflicts:
	hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/namenode/ha/TestBootstrapStandby.java

Change-Id: Iffa62547483c1d9dc82a196d5e8d0856b397217d
This commit is contained in:
Zhe Zhang 2015-11-02 10:03:39 -08:00
parent c1d7b26e94
commit a47c4e7819
2 changed files with 75 additions and 22 deletions

View File

@ -1364,6 +1364,9 @@ Release 2.8.0 - UNRELEASED
HDFS-9343. Empty caller context considered invalid. (Mingliang Liu via
Arpit Agarwal)
HDFS-9329. TestBootstrapStandby#testRateThrottling is flaky because fsimage
size is smaller than IO buffer size. (zhz)
Release 2.7.2 - UNRELEASED
INCOMPATIBLE CHANGES

View File

@ -25,13 +25,16 @@ import java.io.File;
import java.io.IOException;
import java.net.URI;
import java.util.concurrent.TimeoutException;
import java.util.concurrent.atomic.AtomicBoolean;
import com.google.common.base.Supplier;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileUtil;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.hdfs.DFSConfigKeys;
import org.apache.hadoop.hdfs.DFSUtilClient;
import org.apache.hadoop.hdfs.MiniDFSCluster;
import org.apache.hadoop.hdfs.MiniDFSNNTopology;
import org.apache.hadoop.hdfs.server.namenode.CheckpointSignature;
@ -97,6 +100,8 @@ public class TestBootstrapStandby {
"storage directory does not exist or is not accessible",
ioe);
}
int expectedCheckpointTxId = (int)NameNodeAdapter.getNamesystem(nn0)
.getFSImage().getMostRecentCheckpointTxId();
int rc = BootstrapStandby.run(
new String[]{"-nonInteractive"},
@ -105,7 +110,7 @@ public class TestBootstrapStandby {
// Should have copied over the namespace from the active
FSImageTestUtil.assertNNHasCheckpoints(cluster, 1,
ImmutableList.of(0));
ImmutableList.of(expectedCheckpointTxId));
FSImageTestUtil.assertNNFilesMatch(cluster);
// We should now be able to start the standby successfully.
@ -214,7 +219,7 @@ public class TestBootstrapStandby {
* {@link DFSConfigKeys#DFS_IMAGE_TRANSFER_BOOTSTRAP_STANDBY_RATE_KEY}
* created by HDFS-8808.
*/
@Test
@Test(timeout=30000)
public void testRateThrottling() throws Exception {
cluster.getConfiguration(0).setLong(
DFSConfigKeys.DFS_IMAGE_TRANSFER_RATE_KEY, 1);
@ -222,23 +227,29 @@ public class TestBootstrapStandby {
cluster.waitActive();
nn0 = cluster.getNameNode(0);
cluster.transitionToActive(0);
// Each edit has at least 1 byte. So the lowRate definitely should cause
// a timeout, if enforced. If lowRate is not enforced, any reasonable test
// machine should at least download an image with 5 edits in 5 seconds.
for (int i = 0; i < 5; i++) {
nn0.getRpcServer().rollEditLog();
}
int timeOut = updatePrimaryNNAndGetTimeout();
// A very low DFS_IMAGE_TRANSFER_RATE_KEY value won't affect bootstrapping
final AtomicBoolean bootStrapped = new AtomicBoolean(false);
new Thread(
new Runnable() {
@Override
public void run() {
try {
testSuccessfulBaseCase();
bootStrapped.set(true);
} catch (Exception e) {
fail(e.getMessage());
}
}
}
).start();
GenericTestUtils.waitFor(new Supplier<Boolean>() {
public Boolean get() {
try {
testSuccessfulBaseCase();
return true;
} catch (Exception e) {
return false;
}
return bootStrapped.get();
}
}, 500, 5000);
}, 50, timeOut);
shutdownCluster();
setupCluster();
@ -250,22 +261,61 @@ public class TestBootstrapStandby {
cluster.transitionToActive(0);
// A very low DFS_IMAGE_TRANSFER_BOOTSTRAP_STANDBY_RATE_KEY value should
// cause timeout
timeOut = updatePrimaryNNAndGetTimeout();
bootStrapped.set(false);
new Thread(
new Runnable() {
@Override
public void run() {
try {
testSuccessfulBaseCase();
bootStrapped.set(true);
} catch (Exception e) {
LOG.info(e.getMessage());
}
}
}
).start();
try {
GenericTestUtils.waitFor(new Supplier<Boolean>() {
public Boolean get() {
try {
testSuccessfulBaseCase();
return true;
} catch (Exception e) {
return false;
}
return bootStrapped.get();
}
}, 500, 5000);
}, 50, timeOut);
fail("Did not timeout");
} catch (TimeoutException e) {
LOG.info("Encountered expected timeout.");
}
}
/**
* Add enough content to the primary NN's fsimage so that it's larger than
* the IO transfer buffer size of bootstrapping. The return the correct
* timeout duration.
*/
private int updatePrimaryNNAndGetTimeout() throws IOException{
// Any reasonable test machine should be able to transfer 1 byte per MS
// (which is ~1K/s)
final int minXferRatePerMS = 1;
int imageXferBufferSize = DFSUtilClient.getIoFileBufferSize(
new Configuration());
File imageFile = null;
int dirIdx = 0;
while (imageFile == null || imageFile.length() < imageXferBufferSize) {
for (int i = 0; i < 5; i++) {
cluster.getFileSystem(0).mkdirs(new Path("/foo" + dirIdx++));
}
nn0.getRpcServer().rollEditLog();
NameNodeAdapter.enterSafeMode(nn0, false);
NameNodeAdapter.saveNamespace(nn0);
NameNodeAdapter.leaveSafeMode(nn0);
imageFile = FSImageTestUtil.findLatestImageFile(FSImageTestUtil
.getFSImage(nn0).getStorage().getStorageDir(0));
}
return (int)(imageFile.length() / minXferRatePerMS) + 1;
}
private void removeStandbyNameDirs() {
for (URI u : cluster.getNameDirs(1)) {
assertTrue(u.getScheme().equals("file"));