HBASE-12761 On region jump ClientScanners should get next row start key instead of a skip.

Signed-off-by: stack <stack@apache.org>

Conflicts:
	hbase-client/src/main/java/org/apache/hadoop/hbase/client/ScannerCallableWithReplicas.java
This commit is contained in:
Jurriaan Mous 2014-12-28 13:53:03 +01:00 committed by stack
parent 2ff16e532b
commit 26749c347d
3 changed files with 33 additions and 49 deletions

View File

@ -45,6 +45,8 @@ import org.apache.hadoop.hbase.util.Bytes;
import com.google.common.annotations.VisibleForTesting;
import static org.apache.hadoop.hbase.client.ReversedClientScanner.createClosestRowBefore;
/**
* Implements the scanner interface for the HBase client.
* If there are multiple regions in a table, this scanner will iterate
@ -92,7 +94,8 @@ public class ClientScanner extends AbstractClientScanner {
*/
public ClientScanner(final Configuration conf, final Scan scan, final TableName tableName,
ClusterConnection connection, RpcRetryingCallerFactory rpcFactory,
RpcControllerFactory controllerFactory, ExecutorService pool, int primaryOperationTimeout) throws IOException {
RpcControllerFactory controllerFactory, ExecutorService pool, int primaryOperationTimeout)
throws IOException {
if (LOG.isTraceEnabled()) {
LOG.trace("Scan table=" + tableName
+ ", startRow=" + Bytes.toStringBinary(scan.getStartRow()));
@ -229,7 +232,7 @@ public class ClientScanner extends AbstractClientScanner {
// Close the previous scanner if it's open
if (this.callable != null) {
this.callable.setClose();
call(scan, callable, caller, scannerTimeout);
call(callable, caller, scannerTimeout);
this.callable = null;
}
@ -266,7 +269,7 @@ public class ClientScanner extends AbstractClientScanner {
callable = getScannerCallable(localStartKey, nbRows);
// Open a scanner on the region server starting at the
// beginning of the region
call(scan, callable, caller, scannerTimeout);
call(callable, caller, scannerTimeout);
this.currentRegion = callable.getHRegionInfo();
if (this.scanMetrics != null) {
this.scanMetrics.countOfRegions.incrementAndGet();
@ -283,7 +286,7 @@ public class ClientScanner extends AbstractClientScanner {
return callable.isAnyRPCcancelled();
}
static Result[] call(Scan scan, ScannerCallableWithReplicas callable,
static Result[] call(ScannerCallableWithReplicas callable,
RpcRetryingCaller<Result[]> caller, int scannerTimeout)
throws IOException, RuntimeException {
if (Thread.interrupted()) {
@ -310,12 +313,12 @@ public class ClientScanner extends AbstractClientScanner {
/**
* Publish the scan metrics. For now, we use scan.setAttribute to pass the metrics back to the
* application or TableInputFormat.Later, we could push it to other systems. We don't use metrics
* framework because it doesn't support multi-instances of the same metrics on the same machine;
* for scan/map reduce scenarios, we will have multiple scans running at the same time.
* application or TableInputFormat.Later, we could push it to other systems. We don't use
* metrics framework because it doesn't support multi-instances of the same metrics on the same
* machine; for scan/map reduce scenarios, we will have multiple scans running at the same time.
*
* By default, scan metrics are disabled; if the application wants to collect them, this behavior
* can be turned on by calling calling:
* By default, scan metrics are disabled; if the application wants to collect them, this
* behavior can be turned on by calling calling:
*
* scan.setAttribute(SCAN_ATTRIBUTES_METRICS_ENABLE, Bytes.toBytes(Boolean.TRUE))
*/
@ -343,39 +346,13 @@ public class ClientScanner extends AbstractClientScanner {
callable.setCaching(this.caching);
// This flag is set when we want to skip the result returned. We do
// this when we reset scanner because it split under us.
boolean skipFirst = false;
boolean retryAfterOutOfOrderException = true;
do {
try {
if (skipFirst) {
// Skip only the first row (which was the last row of the last
// already-processed batch).
callable.setCaching(1);
values = call(scan, callable, caller, scannerTimeout);
// When the replica switch happens, we need to do certain operations
// again. The scannercallable will openScanner with the right startkey
// but we need to pick up from there. Bypass the rest of the loop
// and let the catch-up happen in the beginning of the loop as it
// happens for the cases where we see exceptions. Since only openScanner
// would have happened, values would be null
if (values == null && callable.switchedToADifferentReplica()) {
if (this.lastResult != null) { //only skip if there was something read earlier
skipFirst = true;
}
this.currentRegion = callable.getHRegionInfo();
continue;
}
callable.setCaching(this.caching);
skipFirst = false;
}
// Server returns a null values if scanning is to stop. Else,
// returns an empty array if scanning is to go on and we've just
// exhausted current region.
values = call(scan, callable, caller, scannerTimeout);
if (skipFirst && values != null && values.length == 1) {
skipFirst = false; // Already skipped, unset it before scanning again
values = call(scan, callable, caller, scannerTimeout);
}
values = call(callable, caller, scannerTimeout);
// When the replica switch happens, we need to do certain operations
// again. The callable will openScanner with the right startkey
// but we need to pick up from there. Bypass the rest of the loop
@ -383,9 +360,6 @@ public class ClientScanner extends AbstractClientScanner {
// happens for the cases where we see exceptions. Since only openScanner
// would have happened, values would be null
if (values == null && callable.switchedToADifferentReplica()) {
if (this.lastResult != null) { //only skip if there was something read earlier
skipFirst = true;
}
this.currentRegion = callable.getHRegionInfo();
continue;
}
@ -428,11 +402,11 @@ public class ClientScanner extends AbstractClientScanner {
// scanner starts at the correct row. Otherwise we may see previously
// returned rows again.
// (ScannerCallable by now has "relocated" the correct region)
this.scan.setStartRow(this.lastResult.getRow());
// Skip first row returned. We already let it out on previous
// invocation.
skipFirst = true;
if(scan.isReversed()){
scan.setStartRow(createClosestRowBefore(lastResult.getRow()));
}else {
scan.setStartRow(Bytes.add(lastResult.getRow(), new byte[1]));
}
}
if (e instanceof OutOfOrderScannerNextException) {
if (retryAfterOutOfOrderException) {
@ -452,7 +426,7 @@ public class ClientScanner extends AbstractClientScanner {
continue;
}
long currentTime = System.currentTimeMillis();
if (this.scanMetrics != null ) {
if (this.scanMetrics != null) {
this.scanMetrics.sumOfMillisSecBetweenNexts.addAndGet(currentTime-lastNext);
}
lastNext = currentTime;
@ -487,7 +461,7 @@ public class ClientScanner extends AbstractClientScanner {
if (callable != null) {
callable.setClose();
try {
call(scan, callable, caller, scannerTimeout);
call(callable, caller, scannerTimeout);
} catch (UnknownScannerException e) {
// We used to catch this error, interpret, and rethrow. However, we
// have since decided that it's not nice for a scanner's close to

View File

@ -57,7 +57,8 @@ public class ReversedClientScanner extends ClientScanner {
TableName tableName, ClusterConnection connection,
RpcRetryingCallerFactory rpcFactory, RpcControllerFactory controllerFactory,
ExecutorService pool, int primaryOperationTimeout) throws IOException {
super(conf, scan, tableName, connection, rpcFactory, controllerFactory, pool, primaryOperationTimeout);
super(conf, scan, tableName, connection, rpcFactory, controllerFactory, pool,
primaryOperationTimeout);
}
@Override
@ -166,7 +167,7 @@ public class ReversedClientScanner extends ClientScanner {
* @param row
* @return a new byte array which is the closest front row of the specified one
*/
protected byte[] createClosestRowBefore(byte[] row) {
protected static byte[] createClosestRowBefore(byte[] row) {
if (row == null) {
throw new IllegalArgumentException("The passed row is empty");
}

View File

@ -39,9 +39,13 @@ import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.hbase.HRegionInfo;
import org.apache.hadoop.hbase.RegionLocations;
import org.apache.hadoop.hbase.TableName;
import org.apache.hadoop.hbase.util.Bytes;
import org.apache.hadoop.hbase.util.Pair;
import com.google.common.annotations.VisibleForTesting;
import static org.apache.hadoop.hbase.client.ReversedClientScanner.createClosestRowBefore;
/**
* This class has the logic for handling scanners for regions with and without replicas.
* 1. A scan is attempted on the default (primary) region
@ -272,8 +276,13 @@ class ScannerCallableWithReplicas implements RetryingCallable<Result[]> {
continue; //this was already scheduled earlier
}
ScannerCallable s = currentScannerCallable.getScannerCallableForReplica(id);
if (this.lastResult != null) {
s.getScan().setStartRow(this.lastResult.getRow());
if(s.getScan().isReversed()){
s.getScan().setStartRow(createClosestRowBefore(this.lastResult.getRow()));
}else {
s.getScan().setStartRow(Bytes.add(this.lastResult.getRow(), new byte[1]));
}
}
outstandingCallables.add(s);
RetryingRPC retryingOnReplica = new RetryingRPC(s);