HBASE-27794: Tooling for parsing/reading the prefetch files list file (apache#5468)

Kota-SH · wchevreuil · commit 497ffe248402 · 2023-11-02T22:24:21.000Z
Signed-off-by: Wellingt
diff --git a/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Admin.java b/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Admin.java
@@ -2629,4 +2629,9 @@ List<LogEntry> getLogEntries(Set<ServerName> serverNames, String logType, Server
    * Flush master local region
    */
   void flushMasterStore() throws IOException;
+
+  /**
+   * Get the list of cached files
+   */
+  List<String> getCachedFilesList(ServerName serverName) throws IOException;
 }
diff --git a/hbase-client/src/main/java/org/apache/hadoop/hbase/client/AdminOverAsyncAdmin.java b/hbase-client/src/main/java/org/apache/hadoop/hbase/client/AdminOverAsyncAdmin.java
@@ -1125,4 +1125,9 @@ public List<LogEntry> getLogEntries(Set<ServerName> serverNames, String logType,
   public void flushMasterStore() throws IOException {
     get(admin.flushMasterStore());
   }
+
+  @Override
+  public List<String> getCachedFilesList(ServerName serverName) throws IOException {
+    return get(admin.getCachedFilesList(serverName));
+  }
 }
diff --git a/hbase-client/src/main/java/org/apache/hadoop/hbase/client/AsyncAdmin.java b/hbase-client/src/main/java/org/apache/hadoop/hbase/client/AsyncAdmin.java
@@ -1843,4 +1843,9 @@ CompletableFuture<List<LogEntry>> getLogEntries(Set<ServerName> serverNames, Str
    * Flush master local region
    */
   CompletableFuture<Void> flushMasterStore();
+
+  /**
+   * Get the list of cached files
+   */
+  CompletableFuture<List<String>> getCachedFilesList(ServerName serverName);
 }
diff --git a/hbase-client/src/main/java/org/apache/hadoop/hbase/client/AsyncHBaseAdmin.java b/hbase-client/src/main/java/org/apache/hadoop/hbase/client/AsyncHBaseAdmin.java
@@ -995,4 +995,9 @@ public CompletableFuture<List<LogEntry>> getLogEntries(Set<ServerName> serverNam
   public CompletableFuture<Void> flushMasterStore() {
     return wrap(rawAdmin.flushMasterStore());
   }
+
+  @Override
+  public CompletableFuture<List<String>> getCachedFilesList(ServerName serverName) {
+    return wrap(rawAdmin.getCachedFilesList(serverName));
+  }
 }
diff --git a/hbase-client/src/main/java/org/apache/hadoop/hbase/client/RawAsyncHBaseAdmin.java b/hbase-client/src/main/java/org/apache/hadoop/hbase/client/RawAsyncHBaseAdmin.java
@@ -132,6 +132,8 @@
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.CompactionSwitchResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.FlushRegionRequest;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.FlushRegionResponse;
+import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetCachedFilesListRequest;
+import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetCachedFilesListResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetOnlineRegionRequest;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetOnlineRegionResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetRegionInfoRequest;
@@ -4519,4 +4521,15 @@ Void> call(controller, stub, request.build(),
           (s, c, req, done) -> s.flushMasterStore(c, req, done), resp -> null))
       .call();
   }
+
+  @Override
+  public CompletableFuture<List<String>> getCachedFilesList(ServerName serverName) {
+    GetCachedFilesListRequest.Builder request = GetCachedFilesListRequest.newBuilder();
+    return this.<List<String>> newAdminCaller()
+      .action((controller, stub) -> this.<GetCachedFilesListRequest, GetCachedFilesListResponse,
+        List<String>> adminCall(controller, stub, request.build(),
+          (s, c, req, done) -> s.getCachedFilesList(c, req, done),
+          resp -> resp.getCachedFilesList()))
+      .serverName(serverName).call();
+  }
 }
diff --git a/hbase-client/src/main/java/org/apache/hadoop/hbase/shaded/protobuf/ProtobufUtil.java b/hbase-client/src/main/java/org/apache/hadoop/hbase/shaded/protobuf/ProtobufUtil.java
@@ -153,6 +153,8 @@
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.AdminService;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.ClearSlowLogResponses;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.CloseRegionRequest;
+import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetCachedFilesListRequest;
+import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetCachedFilesListResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetOnlineRegionRequest;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetOnlineRegionResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetRegionInfoRequest;
@@ -1780,6 +1782,21 @@ public static List<org.apache.hadoop.hbase.client.RegionInfo> getOnlineRegions(
     return getRegionInfos(response);
   }
 
+  /**
+   * Get the list of cached files
+   */
+  public static List<String> getCachedFilesList(final RpcController controller,
+    final AdminService.BlockingInterface admin) throws IOException {
+    GetCachedFilesListRequest request = GetCachedFilesListRequest.newBuilder().build();
+    GetCachedFilesListResponse response = null;
+    try {
+      response = admin.getCachedFilesList(controller, request);
+    } catch (ServiceException se) {
+      throw getRemoteException(se);
+    }
+    return new ArrayList<>(response.getCachedFilesList());
+  }
+
   /**
    * Get the list of region info from a GetOnlineRegionResponse
    * @param proto the GetOnlineRegionResponse
diff --git a/hbase-protocol-shaded/src/main/protobuf/server/region/Admin.proto b/hbase-protocol-shaded/src/main/protobuf/server/region/Admin.proto
@@ -283,6 +283,13 @@ message ExecuteProceduresRequest {
 message ExecuteProceduresResponse {
 }
 
+message GetCachedFilesListRequest {
+}
+
+message GetCachedFilesListResponse {
+  repeated string cached_files = 1;
+}
+
 /**
  * Slow/Large log (LogRequest) use-case specific RPC request. This request payload will be
  * converted in bytes and sent to generic RPC API: GetLogEntries
@@ -405,4 +412,7 @@ service AdminService {
   rpc GetLogEntries(LogRequest)
     returns(LogEntry);
 
+  rpc GetCachedFilesList(GetCachedFilesListRequest)
+    returns(GetCachedFilesListResponse);
+
 }
diff --git a/hbase-server/src/main/java/org/apache/hadoop/hbase/io/hfile/BlockCache.java b/hbase-server/src/main/java/org/apache/hadoop/hbase/io/hfile/BlockCache.java
@@ -18,6 +18,9 @@
 package org.apache.hadoop.hbase.io.hfile;
 
 import java.util.Iterator;
+import java.util.Map;
+import java.util.Optional;
+import org.apache.hadoop.hbase.util.Pair;
 import org.apache.yetus.audience.InterfaceAudience;
 
 /**
@@ -161,4 +164,11 @@ default Cacheable getBlock(BlockCacheKey cacheKey, boolean caching, boolean repe
   default boolean isMetaBlock(BlockType blockType) {
     return blockType != null && blockType.getCategory() != BlockType.BlockCategory.DATA;
   }
+
+  /**
+   * Returns the list of fully cached files
+   */
+  default Optional<Map<String, Pair<String, Long>>> getFullyCachedFiles() {
+    return Optional.empty();
+  }
 }
diff --git a/hbase-server/src/main/java/org/apache/hadoop/hbase/io/hfile/CombinedBlockCache.java b/hbase-server/src/main/java/org/apache/hadoop/hbase/io/hfile/CombinedBlockCache.java
@@ -18,8 +18,11 @@
 package org.apache.hadoop.hbase.io.hfile;
 
 import java.util.Iterator;
+import java.util.Map;
+import java.util.Optional;
 import org.apache.hadoop.hbase.io.HeapSize;
 import org.apache.hadoop.hbase.io.hfile.bucket.BucketCache;
+import org.apache.hadoop.hbase.util.Pair;
 import org.apache.yetus.audience.InterfaceAudience;
 
 /**
@@ -382,6 +385,14 @@ public BlockCache[] getBlockCaches() {
     return new BlockCache[] { this.l1Cache, this.l2Cache };
   }
 
+  /**
+   * Returns the list of fully cached files
+   */
+  @Override
+  public Optional<Map<String, Pair<String, Long>>> getFullyCachedFiles() {
+    return this.l2Cache.getFullyCachedFiles();
+  }
+
   @Override
   public void setMaxSize(long size) {
     this.l1Cache.setMaxSize(size);
diff --git a/hbase-server/src/main/java/org/apache/hadoop/hbase/io/hfile/HFilePreadReader.java b/hbase-server/src/main/java/org/apache/hadoop/hbase/io/hfile/HFilePreadReader.java
@@ -42,8 +42,9 @@ public HFilePreadReader(ReaderContext context, HFileInfo fileInfo, CacheConfig c
     final MutableBoolean fileAlreadyCached = new MutableBoolean(false);
     Optional<BucketCache> bucketCacheOptional =
       BucketCache.getBucketCacheFromCacheConfig(cacheConf);
-    bucketCacheOptional.ifPresent(bc -> fileAlreadyCached
-      .setValue(bc.getFullyCachedFiles().get(path.getName()) == null ? false : true));
+    bucketCacheOptional.flatMap(BucketCache::getFullyCachedFiles).ifPresent(fcf -> {
+      fileAlreadyCached.setValue(fcf.get(path.getName()) == null ? false : true);
+    });
     // Prefetch file blocks upon open if requested
     if (
       cacheConf.shouldPrefetchOnOpen() && cacheIfCompactionsOff()
diff --git a/hbase-server/src/main/java/org/apache/hadoop/hbase/io/hfile/bucket/BucketCache.java b/hbase-server/src/main/java/org/apache/hadoop/hbase/io/hfile/bucket/BucketCache.java
@@ -2002,8 +2002,9 @@ public AtomicBoolean getBackingMapValidated() {
     return backingMapValidated;
   }
 
-  public Map<String, Pair<String, Long>> getFullyCachedFiles() {
-    return fullyCachedFiles;
+  @Override
+  public Optional<Map<String, Pair<String, Long>>> getFullyCachedFiles() {
+    return Optional.of(fullyCachedFiles);
   }
 
   public static Optional<BucketCache> getBucketCacheFromCacheConfig(CacheConfig cacheConf) {
diff --git a/hbase-server/src/main/java/org/apache/hadoop/hbase/master/MasterRpcServices.java b/hbase-server/src/main/java/org/apache/hadoop/hbase/master/MasterRpcServices.java
@@ -164,6 +164,8 @@
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.ExecuteProceduresResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.FlushRegionRequest;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.FlushRegionResponse;
+import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetCachedFilesListRequest;
+import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetCachedFilesListResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetOnlineRegionRequest;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetOnlineRegionResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetRegionInfoRequest;
@@ -3575,6 +3577,12 @@ public ExecuteProceduresResponse executeProcedures(RpcController controller,
     throw new ServiceException(new DoNotRetryIOException("Unsupported method on master"));
   }
 
+  @Override
+  public GetCachedFilesListResponse getCachedFilesList(RpcController controller,
+    GetCachedFilesListRequest request) throws ServiceException {
+    throw new ServiceException(new DoNotRetryIOException("Unsupported method on master"));
+  }
+
   @Override
   public GetLiveRegionServersResponse getLiveRegionServers(RpcController controller,
     GetLiveRegionServersRequest request) throws ServiceException {
diff --git a/hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/RSRpcServices.java b/hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/RSRpcServices.java
@@ -169,6 +169,8 @@
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.ExecuteProceduresResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.FlushRegionRequest;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.FlushRegionResponse;
+import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetCachedFilesListRequest;
+import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetCachedFilesListResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetOnlineRegionRequest;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetOnlineRegionResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetRegionInfoRequest;
@@ -3933,4 +3935,15 @@ public void onConfigurationChange(Configuration conf) {
     super.onConfigurationChange(conf);
     setReloadableGuardrails(conf);
   }
+
+  @Override
+  public GetCachedFilesListResponse getCachedFilesList(RpcController controller,
+    GetCachedFilesListRequest request) throws ServiceException {
+    GetCachedFilesListResponse.Builder responseBuilder = GetCachedFilesListResponse.newBuilder();
+    List<String> fullyCachedFiles = new ArrayList<>();
+    server.getBlockCache().flatMap(BlockCache::getFullyCachedFiles).ifPresent(fcf -> {
+      fullyCachedFiles.addAll(fcf.keySet());
+    });
+    return responseBuilder.addAllCachedFiles(fullyCachedFiles).build();
+  }
 }
diff --git a/hbase-server/src/test/java/org/apache/hadoop/hbase/io/hfile/TestPrefetchRSClose.java b/hbase-server/src/test/java/org/apache/hadoop/hbase/io/hfile/TestPrefetchRSClose.java
@@ -23,18 +23,22 @@
 import static org.junit.Assert.assertTrue;
 
 import java.io.File;
+import java.util.List;
 import org.apache.hadoop.conf.Configuration;
 import org.apache.hadoop.fs.Path;
 import org.apache.hadoop.hbase.HBaseClassTestRule;
 import org.apache.hadoop.hbase.HBaseTestingUtil;
 import org.apache.hadoop.hbase.SingleProcessHBaseCluster;
 import org.apache.hadoop.hbase.StartTestingClusterOption;
 import org.apache.hadoop.hbase.TableName;
+import org.apache.hadoop.hbase.client.Admin;
 import org.apache.hadoop.hbase.client.ColumnFamilyDescriptorBuilder;
 import org.apache.hadoop.hbase.client.Put;
 import org.apache.hadoop.hbase.client.Table;
 import org.apache.hadoop.hbase.client.TableDescriptor;
 import org.apache.hadoop.hbase.client.TableDescriptorBuilder;
+import org.apache.hadoop.hbase.regionserver.HRegionServer;
+import org.apache.hadoop.hbase.regionserver.HStoreFile;
 import org.apache.hadoop.hbase.testclassification.IOTests;
 import org.apache.hadoop.hbase.testclassification.LargeTests;
 import org.apache.hadoop.hbase.util.Bytes;
@@ -112,7 +116,17 @@ public void testPrefetchPersistence() throws Exception {
 
     // Default interval for cache persistence is 1000ms. So after 1000ms, both the persistence files
     // should exist.
-    assertTrue(new File(testDir + "/bucket.persistence").exists());
+
+    HRegionServer regionServingRS = cluster.getRegionServer(1).getRegions(tableName).size() == 1
+      ? cluster.getRegionServer(1)
+      : cluster.getRegionServer(0);
+
+    Admin admin = TEST_UTIL.getAdmin();
+    List<String> cachedFilesList = admin.getCachedFilesList(regionServingRS.getServerName());
+    assertEquals(1, cachedFilesList.size());
+    for (HStoreFile h : regionServingRS.getRegions().get(0).getStores().get(0).getStorefiles()) {
+      assertTrue(cachedFilesList.contains(h.getPath().getName()));
+    }
 
     // Stop the RS
     cluster.stopRegionServer(0);
diff --git a/hbase-server/src/test/java/org/apache/hadoop/hbase/io/hfile/TestPrefetchWithBucketCache.java b/hbase-server/src/test/java/org/apache/hadoop/hbase/io/hfile/TestPrefetchWithBucketCache.java
@@ -123,7 +123,7 @@ public void testPrefetchDoesntOverwork() throws Exception {
     BlockCacheKey key = snapshot.keySet().stream().findFirst().get();
     LOG.debug("removing block {}", key);
     bc.getBackingMap().remove(key);
-    bc.getFullyCachedFiles().remove(storeFile.getName());
+    bc.getFullyCachedFiles().ifPresent(fcf -> fcf.remove(storeFile.getName()));
     assertTrue(snapshot.size() > bc.getBackingMap().size());
     LOG.debug("Third read should prefetch again, as we removed one block for the file.");
     readStoreFile(storeFile);
diff --git a/hbase-server/src/test/java/org/apache/hadoop/hbase/master/MockRegionServer.java b/hbase-server/src/test/java/org/apache/hadoop/hbase/master/MockRegionServer.java
@@ -97,6 +97,8 @@
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.ExecuteProceduresResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.FlushRegionRequest;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.FlushRegionResponse;
+import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetCachedFilesListRequest;
+import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetCachedFilesListResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetOnlineRegionRequest;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetOnlineRegionResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.AdminProtos.GetRegionInfoRequest;
@@ -452,6 +454,12 @@ public GetOnlineRegionResponse getOnlineRegion(RpcController controller,
     return null;
   }
 
+  @Override
+  public GetCachedFilesListResponse getCachedFilesList(RpcController controller,
+    GetCachedFilesListRequest request) throws ServiceException {
+    return null;
+  }
+
   @Override
   public List<HRegion> getRegions() {
     return null;
diff --git a/hbase-server/src/test/java/org/apache/hadoop/hbase/rsgroup/VerifyingRSGroupAdmin.java b/hbase-server/src/test/java/org/apache/hadoop/hbase/rsgroup/VerifyingRSGroupAdmin.java
@@ -969,6 +969,11 @@ public void flushMasterStore() throws IOException {
     admin.flushMasterStore();
   }
 
+  @Override
+  public List<String> getCachedFilesList(ServerName serverName) throws IOException {
+    return admin.getCachedFilesList(serverName);
+  }
+
   @Override
   public boolean replicationPeerModificationSwitch(boolean on, boolean drainProcedures)
     throws IOException {
diff --git a/hbase-thrift/src/main/java/org/apache/hadoop/hbase/thrift2/client/ThriftAdmin.java b/hbase-thrift/src/main/java/org/apache/hadoop/hbase/thrift2/client/ThriftAdmin.java
@@ -1345,6 +1345,11 @@ public void flushMasterStore() throws IOException {
     throw new NotImplementedException("flushMasterStore not supported in ThriftAdmin");
   }
 
+  @Override
+  public List<String> getCachedFilesList(ServerName serverName) throws IOException {
+    throw new NotImplementedException("getCachedFilesList not supported in ThriftAdmin");
+  }
+
   @Override
   public boolean replicationPeerModificationSwitch(boolean on, boolean drainProcedures)
     throws IOException {

Original file line number	Diff line number	Diff line change
`@@ -1125,4 +1125,9 @@ public List<LogEntry> getLogEntries(Set<ServerName> serverNames, String logType,`
`1125`	`1125`	`public void flushMasterStore() throws IOException {`
`1126`	`1126`	`get(admin.flushMasterStore());`
`1127`	`1127`	`}`
	`1128`	`+`
	`1129`	`+ @Override`
	`1130`	`+ public List<String> getCachedFilesList(ServerName serverName) throws IOException {`
	`1131`	`+ return get(admin.getCachedFilesList(serverName));`
	`1132`	`+ }`
`1128`	`1133`	`}`
Original file line number	Diff line number	Diff line change
`@@ -995,4 +995,9 @@ public CompletableFuture<List<LogEntry>> getLogEntries(Set<ServerName> serverNam`
`995`	`995`	`public CompletableFuture<Void> flushMasterStore() {`
`996`	`996`	`return wrap(rawAdmin.flushMasterStore());`
`997`	`997`	`}`
	`998`	`+`
	`999`	`+ @Override`
	`1000`	`+ public CompletableFuture<List<String>> getCachedFilesList(ServerName serverName) {`
	`1001`	`+ return wrap(rawAdmin.getCachedFilesList(serverName));`
	`1002`	`+ }`
`998`	`1003`	`}`