HLog writer can do 1-2 sync operations after lease has been recovered for split process.

git-svn-id: https://svn.apache.org/repos/asf/hbase/trunk@1403977 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Michael Stack 2012-10-31 04:48:37 +00:00
parent b31e78a424
commit fb983be9e3
1 changed files with 75 additions and 64 deletions

View File

@ -85,8 +85,7 @@ public class TestHLogSplit {
private Configuration conf;
private FileSystem fs;
private final static HBaseTestingUtility
TEST_UTIL = new HBaseTestingUtility();
private final static HBaseTestingUtility TEST_UTIL = new HBaseTestingUtility();
private static final Path hbaseDir = new Path("/hbase");
@ -156,6 +155,46 @@ public class TestHLogSplit {
public void tearDown() throws Exception {
}
/**
* Simulates splitting a WAL out from under a regionserver that is still trying to write it. Ensures we do not
* lose edits.
* @throws IOException
*/
@Test
public void testLogCannotBeWrittenOnceParsed() throws IOException {
AtomicLong counter = new AtomicLong(0);
AtomicBoolean stop = new AtomicBoolean(false);
// Index of the WAL we want to keep open. generateHLogs will leave open the WAL whose index we supply here.
int walToKeepOpen = 9;
// The below method writes NUM_WRITERS files each with ENTRIES entries it for a total of NUM_WRITERS * ENTRIES added
// per column family in the region.
generateHLogs(walToKeepOpen);
fs.initialize(fs.getUri(), conf);
String julietRegion = regions.get(0);
// This WAL should be open still
HLog.Writer stillOpenWAL = writer[walToKeepOpen];
// Thread that will keep writing the WAL across the WAL splitting below. Before we start writing, make sure the
// counter has the edits that were written above in generateHLogs.
counter.addAndGet(ENTRIES * NUM_WRITERS);
Thread zombie = new ZombieLastLogWriterRegionServer(stillOpenWAL, counter, stop, TABLE_NAME, julietRegion);
try {
zombie.start();
HLogSplitter logSplitter = HLogSplitter.createLogSplitter(conf, hbaseDir, hlogDir, oldLogDir, fs);
logSplitter.splitLog();
// Get the recovered edits file made by the WAL log splitting process above.
Path logfile = getLogForRegion(hbaseDir, TABLE_NAME, julietRegion);
// It's possible that the writer got an error while appending and didn't count an edit
// however the entry will in fact be written to file and split with the rest
long numberOfEditsInRegion = countHLog(logfile, fs, conf);
assertTrue("The log file could have at most 1 extra log entry, but can't have less. Zombie could write " +
counter.get() + " and logfile had only " + numberOfEditsInRegion + " in logfile=" + logfile,
counter.get() == numberOfEditsInRegion || counter.get() + 1 == numberOfEditsInRegion);
} finally {
stop.set(true);
}
}
/**
* @throws IOException
* @see https://issues.apache.org/jira/browse/HBASE-3020
@ -478,11 +517,11 @@ public class TestHLogSplit {
hbaseDir, hlogDir, oldLogDir, fs);
logSplitter.splitLog();
Path originalLog = (fs.listStatus(oldLogDir))[0].getPath();
Path splitLog = getLogForRegion(hbaseDir, TABLE_NAME, REGION);
int actualCount = 0;
HLog.Reader in = HLogFactory.createReader(fs, splitLog, conf);
@SuppressWarnings("unused")
HLog.Entry entry;
while ((entry = in.next()) != null) ++actualCount;
assertEquals(entryCount-1, actualCount);
@ -542,36 +581,7 @@ public class TestHLogSplit {
// hadoop 0.21 throws FNFE whereas hadoop 0.20 returns null
}
}
/* DISABLED for now. TODO: HBASE-2645
@Test
public void testLogCannotBeWrittenOnceParsed() throws IOException {
AtomicLong counter = new AtomicLong(0);
AtomicBoolean stop = new AtomicBoolean(false);
generateHLogs(9);
fs.initialize(fs.getUri(), conf);
Thread zombie = new ZombieLastLogWriterRegionServer(writer[9], counter, stop);
try {
zombie.start();
HLog.splitLog(hbaseDir, hlogDir, oldLogDir, fs, conf);
Path logfile = getLogForRegion(hbaseDir, TABLE_NAME, "juliet");
// It's possible that the writer got an error while appending and didn't count it
// however the entry will in fact be written to file and split with the rest
long numberOfEditsInRegion = countHLog(logfile, fs, conf);
assertTrue("The log file could have at most 1 extra log entry, but " +
"can't have less. Zombie could write "+counter.get() +" and logfile had only"+ numberOfEditsInRegion+" " + logfile, counter.get() == numberOfEditsInRegion ||
counter.get() + 1 == numberOfEditsInRegion);
} finally {
stop.set(true);
}
}
*/
@Test
public void testSplitWillNotTouchLogsIfNewHLogGetsCreatedAfterSplitStarted()
@ -935,10 +945,16 @@ public class TestHLogSplit {
AtomicBoolean stop;
Path log;
HLog.Writer lastLogWriter;
public ZombieLastLogWriterRegionServer(HLog.Writer writer, AtomicLong counter, AtomicBoolean stop) {
private final byte [] tableName;
private final String regionName;
public ZombieLastLogWriterRegionServer(HLog.Writer writer, AtomicLong counter, AtomicBoolean stop,
final byte [] tableName, final String regionName) {
this.stop = stop;
this.editsCount = counter;
this.lastLogWriter = writer;
this.tableName = tableName;
this.regionName = regionName;
}
@Override
@ -947,13 +963,16 @@ public class TestHLogSplit {
return;
}
flushToConsole("starting");
byte [] regionBytes = Bytes.toBytes(this.regionName);
boolean created = false;
while (true) {
try {
String region = "juliet";
fs.mkdirs(new Path(new Path(hbaseDir, region), region));
appendEntry(lastLogWriter, TABLE_NAME, region.getBytes(),
("r" + editsCount).getBytes(), FAMILY, QUALIFIER, VALUE, 0);
if (!created) {
fs.mkdirs(new Path(new Path(hbaseDir, Bytes.toString(this.tableName)), this.regionName));
created = true;
}
appendEntry(lastLogWriter, TABLE_NAME, regionBytes, ("r" + editsCount).getBytes(), regionBytes,
QUALIFIER, VALUE, 0);
lastLogWriter.sync();
editsCount.incrementAndGet();
try {
@ -961,21 +980,16 @@ public class TestHLogSplit {
} catch (InterruptedException e) {
//
}
} catch (IOException ex) {
if (ex instanceof RemoteException) {
flushToConsole("Juliet: got RemoteException " +
ex.getMessage() + " while writing " + (editsCount.get() + 1));
flushToConsole("Juliet: got RemoteException " + ex.getMessage() +
" while writing " + (editsCount.get() + 1));
break;
} else {
assertTrue("Failed to write " + editsCount.get(), false);
}
}
}
}
}
@ -1149,8 +1163,7 @@ public class TestHLogSplit {
regions.add(regionName);
generateHLogs(-1);
final HLog log = HLogFactory.createHLog(fs, regiondir,
regionName, conf);
HLogFactory.createHLog(fs, regiondir, regionName, conf);
HLogSplitter logSplitter = new HLogSplitter(
conf, hbaseDir, hlogDir, oldLogDir, fs, null) {
@ -1210,14 +1223,12 @@ public class TestHLogSplit {
makeRegionDirs(fs, regions);
fs.mkdirs(hlogDir);
for (int i = 0; i < writers; i++) {
writer[i] = HLogFactory.createWriter(fs, new Path(hlogDir, HLOG_FILE_PREFIX + i),
conf);
writer[i] = HLogFactory.createWriter(fs, new Path(hlogDir, HLOG_FILE_PREFIX + i), conf);
for (int j = 0; j < entries; j++) {
int prefix = 0;
for (String region : regions) {
String row_key = region + prefix++ + i + j;
appendEntry(writer[i], TABLE_NAME, region.getBytes(),
row_key.getBytes(), FAMILY, QUALIFIER, VALUE, seq);
appendEntry(writer[i], TABLE_NAME, region.getBytes(), row_key.getBytes(), FAMILY, QUALIFIER, VALUE, seq);
}
}
if (i != leaveOpen) {