[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r312632071
##
File path:
hadoop-ozone/ozone-recon/src/test/java/org/apache/hadoop/ozone/recon/tasks/TestFileSizeCountTask.java
##
@@ -0,0 +1,140 @@
+/**
+ * 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 org.apache.hadoop.ozone.recon.tasks;
+
+import org.apache.hadoop.ozone.om.OMMetadataManager;
+import org.apache.hadoop.ozone.om.OmMetadataManagerImpl;
+import org.apache.hadoop.ozone.om.helpers.OmKeyInfo;
+import org.apache.hadoop.utils.db.TypedTable;
+import org.junit.Test;
+
+import org.junit.runner.RunWith;
+import org.powermock.core.classloader.annotations.PowerMockIgnore;
+import org.powermock.core.classloader.annotations.PrepareForTest;
+import org.powermock.modules.junit4.PowerMockRunner;
+
+import java.io.IOException;
+
+import static org.apache.hadoop.ozone.recon.tasks.
+OMDBUpdateEvent.OMDBUpdateAction.PUT;
+import static org.junit.Assert.assertEquals;
+
+import static org.mockito.ArgumentMatchers.anyLong;
+import static org.mockito.BDDMockito.given;
+import static org.mockito.Mockito.verify;
+import static org.mockito.Mockito.times;
+import static org.powermock.api.mockito.PowerMockito.mock;
+import static org.powermock.api.mockito.PowerMockito.when;
+
+/**
+ * Unit test for Container Key mapper task.
+ */
+@RunWith(PowerMockRunner.class)
+@PowerMockIgnore({"javax.management.*", "javax.net.ssl.*"})
+@PrepareForTest(OmKeyInfo.class)
+
+public class TestFileSizeCountTask {
+ @Test
+ public void testCalculateBinIndex() {
+FileSizeCountTask fileSizeCountTask = mock(FileSizeCountTask.class);
+
+when(fileSizeCountTask.getMaxFileSizeUpperBound()).
+thenReturn(1125899906842624L);// 1 PB
+when(fileSizeCountTask.getOneKB()).thenReturn(1024L);
+when(fileSizeCountTask.getMaxBinSize()).thenReturn(42);
+when(fileSizeCountTask.calculateBinIndex(anyLong())).thenCallRealMethod();
+when(fileSizeCountTask.nextClosestPowerIndexOfTwo(
+anyLong())).thenCallRealMethod();
+
+long fileSize = 1024L;// 1 KB
+int binIndex = fileSizeCountTask.calculateBinIndex(fileSize);
+assertEquals(1, binIndex);
+
+fileSize = 1023L;// 1KB - 1B
+binIndex = fileSizeCountTask.calculateBinIndex(fileSize);
+assertEquals(0, binIndex);
+
+fileSize = 562949953421312L; // 512 TB
+binIndex = fileSizeCountTask.calculateBinIndex(fileSize);
+assertEquals(40, binIndex);
+
+fileSize = 562949953421313L; // (512 TB + 1B)
+binIndex = fileSizeCountTask.calculateBinIndex(fileSize);
+assertEquals(40, binIndex);
+
+fileSize = 562949953421311L; // (512 TB - 1B)
+binIndex = fileSizeCountTask.calculateBinIndex(fileSize);
+assertEquals(39, binIndex);
+
+fileSize = 1125899906842624L; // 1 PB - last (extra) bin
+binIndex = fileSizeCountTask.calculateBinIndex(fileSize);
+assertEquals(41, binIndex);
+
+fileSize = 10L;
+binIndex = fileSizeCountTask.calculateBinIndex(fileSize);
+assertEquals(7, binIndex);
+
+fileSize = 1125899906842623L; // (1 PB - 1B)
+binIndex = fileSizeCountTask.calculateBinIndex(fileSize);
+assertEquals(40, binIndex);
+
+fileSize = 1125899906842624L * 4; // 4 PB - last extra bin
+binIndex = fileSizeCountTask.calculateBinIndex(fileSize);
+assertEquals(41, binIndex);
+
+fileSize = Long.MAX_VALUE;// extra bin
+binIndex = fileSizeCountTask.calculateBinIndex(fileSize);
+assertEquals(41, binIndex);
+ }
+
+ @Test
+ public void testFileCountBySizeReprocess() throws IOException {
Review comment:
Sure.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
-
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r312281064
##
File path:
hadoop-ozone/ozone-recon/src/test/java/org/apache/hadoop/ozone/recon/api/TestUtilizationService.java
##
@@ -62,66 +64,73 @@ public void setUpResultList() {
(long) i));
}
}
+return resultList;
}
@Test
public void testGetFileCounts() throws IOException {
-setUpResultList();
+List resultList = setUpResultList();
utilizationService = mock(UtilizationService.class);
when(utilizationService.getFileCounts()).thenCallRealMethod();
when(utilizationService.getDao()).thenReturn(fileCountBySizeDao);
when(fileCountBySizeDao.findAll()).thenReturn(resultList);
-utilizationService.getFileCounts();
+Response response = utilizationService.getFileCounts();
+// get result list from Response entity
+List responseList =
+(List) response.getEntity();
+
verify(utilizationService, times(1)).getFileCounts();
verify(fileCountBySizeDao, times(1)).findAll();
-assertEquals(maxBinSize, resultList.size());
+FileSizeCountTask fileSizeCountTask = mock(FileSizeCountTask.class);
+when(fileSizeCountTask.getMaxFileSizeUpperBound()).
+thenReturn(1125899906842624L);
+when(fileSizeCountTask.getMaxBinSize()).thenReturn(maxBinSize);
+when(fileSizeCountTask.calculateBinIndex(anyLong())).thenCallRealMethod();
+assertEquals(maxBinSize, responseList.size());
+
long fileSize = 4096L; // 4KB
-int index = findIndex(fileSize);
-long count = resultList.get(index).getCount();
+int index = fileSizeCountTask.calculateBinIndex(fileSize);
+
+long count = responseList.get(index).getCount();
assertEquals(index, count);
fileSize = 1125899906842624L; // 1PB
-index = findIndex(fileSize);
-count = resultList.get(index).getCount();
+index = fileSizeCountTask.calculateBinIndex(fileSize);
+count = responseList.get(index).getCount();
+//last extra bin for files >= 1PB
assertEquals(maxBinSize - 1, index);
assertEquals(index, count);
fileSize = 1025L; // 1 KB + 1B
-index = findIndex(fileSize);
-count = resultList.get(index).getCount(); //last extra bin for files >= 1PB
+index = fileSizeCountTask.calculateBinIndex(fileSize);
+count = responseList.get(index).getCount();
assertEquals(index, count);
fileSize = 25L;
-index = findIndex(fileSize);
-count = resultList.get(index).getCount();
+index = fileSizeCountTask.calculateBinIndex(fileSize);
+count = responseList.get(index).getCount();
assertEquals(index, count);
fileSize = 1125899906842623L; // 1PB - 1B
-index = findIndex(fileSize);
-count = resultList.get(index).getCount();
+index = fileSizeCountTask.calculateBinIndex(fileSize);
+count = responseList.get(index).getCount();
assertEquals(index, count);
fileSize = 1125899906842624L * 4; // 4 PB
-index = findIndex(fileSize);
-count = resultList.get(index).getCount();
+index = fileSizeCountTask.calculateBinIndex(fileSize);
Review comment:
So, what exactly should testGetFileCounts() test?
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
-
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r312281064
##
File path:
hadoop-ozone/ozone-recon/src/test/java/org/apache/hadoop/ozone/recon/api/TestUtilizationService.java
##
@@ -62,66 +64,73 @@ public void setUpResultList() {
(long) i));
}
}
+return resultList;
}
@Test
public void testGetFileCounts() throws IOException {
-setUpResultList();
+List resultList = setUpResultList();
utilizationService = mock(UtilizationService.class);
when(utilizationService.getFileCounts()).thenCallRealMethod();
when(utilizationService.getDao()).thenReturn(fileCountBySizeDao);
when(fileCountBySizeDao.findAll()).thenReturn(resultList);
-utilizationService.getFileCounts();
+Response response = utilizationService.getFileCounts();
+// get result list from Response entity
+List responseList =
+(List) response.getEntity();
+
verify(utilizationService, times(1)).getFileCounts();
verify(fileCountBySizeDao, times(1)).findAll();
-assertEquals(maxBinSize, resultList.size());
+FileSizeCountTask fileSizeCountTask = mock(FileSizeCountTask.class);
+when(fileSizeCountTask.getMaxFileSizeUpperBound()).
+thenReturn(1125899906842624L);
+when(fileSizeCountTask.getMaxBinSize()).thenReturn(maxBinSize);
+when(fileSizeCountTask.calculateBinIndex(anyLong())).thenCallRealMethod();
+assertEquals(maxBinSize, responseList.size());
+
long fileSize = 4096L; // 4KB
-int index = findIndex(fileSize);
-long count = resultList.get(index).getCount();
+int index = fileSizeCountTask.calculateBinIndex(fileSize);
+
+long count = responseList.get(index).getCount();
assertEquals(index, count);
fileSize = 1125899906842624L; // 1PB
-index = findIndex(fileSize);
-count = resultList.get(index).getCount();
+index = fileSizeCountTask.calculateBinIndex(fileSize);
+count = responseList.get(index).getCount();
+//last extra bin for files >= 1PB
assertEquals(maxBinSize - 1, index);
assertEquals(index, count);
fileSize = 1025L; // 1 KB + 1B
-index = findIndex(fileSize);
-count = resultList.get(index).getCount(); //last extra bin for files >= 1PB
+index = fileSizeCountTask.calculateBinIndex(fileSize);
+count = responseList.get(index).getCount();
assertEquals(index, count);
fileSize = 25L;
-index = findIndex(fileSize);
-count = resultList.get(index).getCount();
+index = fileSizeCountTask.calculateBinIndex(fileSize);
+count = responseList.get(index).getCount();
assertEquals(index, count);
fileSize = 1125899906842623L; // 1PB - 1B
-index = findIndex(fileSize);
-count = resultList.get(index).getCount();
+index = fileSizeCountTask.calculateBinIndex(fileSize);
+count = responseList.get(index).getCount();
assertEquals(index, count);
fileSize = 1125899906842624L * 4; // 4 PB
-index = findIndex(fileSize);
-count = resultList.get(index).getCount();
+index = fileSizeCountTask.calculateBinIndex(fileSize);
Review comment:
So, what exactly should testGetFileCounts() test? as part of assertions?
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
-
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r311792914
##
File path:
hadoop-ozone/ozone-recon/src/main/java/org/apache/hadoop/ozone/recon/tasks/FileSizeCountTask.java
##
@@ -0,0 +1,241 @@
+/**
+ * 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 org.apache.hadoop.ozone.recon.tasks;
+
+import com.google.inject.Inject;
+import org.apache.commons.lang3.tuple.ImmutablePair;
+import org.apache.commons.lang3.tuple.Pair;
+import org.apache.hadoop.ozone.om.OMMetadataManager;
+import org.apache.hadoop.ozone.om.helpers.OmKeyInfo;
+import org.apache.hadoop.utils.db.Table;
+import org.apache.hadoop.utils.db.TableIterator;
+import org.hadoop.ozone.recon.schema.tables.daos.FileCountBySizeDao;
+import org.hadoop.ozone.recon.schema.tables.pojos.FileCountBySize;
+import org.jooq.Configuration;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.Iterator;
+import java.util.List;
+
+/**
+ * Class to iterate over the OM DB and store the counts of existing/new
+ * files binned into ranges (1KB, 2Kb..,4MB,.., 1TB,..1PB) to the Recon
+ * fileSize DB.
+ */
+public class FileSizeCountTask extends ReconDBUpdateTask {
+ private static final Logger LOG =
+ LoggerFactory.getLogger(FileSizeCountTask.class);
+
+ private int maxBinSize = -1;
+ private long maxFileSizeUpperBound = 1125899906842624L; // 1 PB
+ private long[] upperBoundCount;
+ private long oneKb = 1024L;
+ private Collection tables = new ArrayList<>();
+ private FileCountBySizeDao fileCountBySizeDao;
+
+ @Inject
+ public FileSizeCountTask(OMMetadataManager omMetadataManager,
+ Configuration sqlConfiguration) {
+super("FileSizeCountTask");
+try {
+ tables.add(omMetadataManager.getKeyTable().getName());
+ fileCountBySizeDao = new FileCountBySizeDao(sqlConfiguration);
+} catch (Exception e) {
+ LOG.error("Unable to fetch Key Table updates ", e);
+}
+upperBoundCount = new long[getMaxBinSize()];
+ }
+
+ protected long getOneKB() {
+return oneKb;
+ }
+
+ protected long getMaxFileSizeUpperBound() {
+return maxFileSizeUpperBound;
+ }
+
+ protected int getMaxBinSize() {
+if (maxBinSize == -1) {
+ // extra bin to add files > 1PB.
+ maxBinSize = calculateBinIndex(maxFileSizeUpperBound) + 1;
+}
+return maxBinSize;
+ }
+
+ /**
+ * Read the Keys from OM snapshot DB and calculate the upper bound of
+ * File Size it belongs to.
+ *
+ * @param omMetadataManager OM Metadata instance.
+ * @return Pair
+ */
+ @Override
+ public Pair reprocess(OMMetadataManager omMetadataManager) {
+LOG.info("Starting a 'reprocess' run of FileSizeCountTask.");
+Table omKeyInfoTable = omMetadataManager.getKeyTable();
+try (TableIterator>
+keyIter = omKeyInfoTable.iterator()) {
+ while (keyIter.hasNext()) {
+Table.KeyValue kv = keyIter.next();
+countFileSize(kv.getValue());
+ }
+} catch (IOException ioEx) {
+ LOG.error("Unable to populate File Size Count in Recon DB. ", ioEx);
+ return new ImmutablePair<>(getTaskName(), false);
+}
+populateFileCountBySizeDB();
+
+LOG.info("Completed a 'reprocess' run of FileSizeCountTask.");
+return new ImmutablePair<>(getTaskName(), true);
+ }
+
+ @Override
+ protected Collection getTaskTables() {
+return tables;
+ }
+
+ void updateCountFromDB() {
+// Read - Write operations to DB are in ascending order
+// of file size upper bounds.
+List resultSet = fileCountBySizeDao.findAll();
+int index = 0;
+if (resultSet != null) {
+ for (FileCountBySize row : resultSet) {
+upperBoundCount[index] = row.getCount();
+index++;
+ }
+}
+ }
+
+ /**
+ * Read the Keys from update events and update the count of files
+ * pertaining to a certain upper bound.
+ *
+ * @param events Update events - PUT/DELETE.
+ * @return Pair
+ */
+ @Override
+ Pair process(OMUpdateEventBatch events) {
+
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r311792886
##
File path:
hadoop-ozone/ozone-recon/src/main/java/org/apache/hadoop/ozone/recon/tasks/FileSizeCountTask.java
##
@@ -0,0 +1,241 @@
+/**
+ * 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 org.apache.hadoop.ozone.recon.tasks;
+
+import com.google.inject.Inject;
+import org.apache.commons.lang3.tuple.ImmutablePair;
+import org.apache.commons.lang3.tuple.Pair;
+import org.apache.hadoop.ozone.om.OMMetadataManager;
+import org.apache.hadoop.ozone.om.helpers.OmKeyInfo;
+import org.apache.hadoop.utils.db.Table;
+import org.apache.hadoop.utils.db.TableIterator;
+import org.hadoop.ozone.recon.schema.tables.daos.FileCountBySizeDao;
+import org.hadoop.ozone.recon.schema.tables.pojos.FileCountBySize;
+import org.jooq.Configuration;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.Iterator;
+import java.util.List;
+
+/**
+ * Class to iterate over the OM DB and store the counts of existing/new
+ * files binned into ranges (1KB, 2Kb..,4MB,.., 1TB,..1PB) to the Recon
+ * fileSize DB.
+ */
+public class FileSizeCountTask extends ReconDBUpdateTask {
+ private static final Logger LOG =
+ LoggerFactory.getLogger(FileSizeCountTask.class);
+
+ private int maxBinSize = -1;
+ private long maxFileSizeUpperBound = 1125899906842624L; // 1 PB
+ private long[] upperBoundCount;
+ private long oneKb = 1024L;
+ private Collection tables = new ArrayList<>();
+ private FileCountBySizeDao fileCountBySizeDao;
+
+ @Inject
+ public FileSizeCountTask(OMMetadataManager omMetadataManager,
+ Configuration sqlConfiguration) {
+super("FileSizeCountTask");
+try {
+ tables.add(omMetadataManager.getKeyTable().getName());
+ fileCountBySizeDao = new FileCountBySizeDao(sqlConfiguration);
+} catch (Exception e) {
+ LOG.error("Unable to fetch Key Table updates ", e);
+}
+upperBoundCount = new long[getMaxBinSize()];
+ }
+
+ protected long getOneKB() {
+return oneKb;
+ }
+
+ protected long getMaxFileSizeUpperBound() {
+return maxFileSizeUpperBound;
+ }
+
+ protected int getMaxBinSize() {
+if (maxBinSize == -1) {
+ // extra bin to add files > 1PB.
+ maxBinSize = calculateBinIndex(maxFileSizeUpperBound) + 1;
+}
+return maxBinSize;
+ }
+
+ /**
+ * Read the Keys from OM snapshot DB and calculate the upper bound of
+ * File Size it belongs to.
+ *
+ * @param omMetadataManager OM Metadata instance.
+ * @return Pair
+ */
+ @Override
+ public Pair reprocess(OMMetadataManager omMetadataManager) {
+LOG.info("Starting a 'reprocess' run of FileSizeCountTask.");
+Table omKeyInfoTable = omMetadataManager.getKeyTable();
+try (TableIterator>
+keyIter = omKeyInfoTable.iterator()) {
+ while (keyIter.hasNext()) {
+Table.KeyValue kv = keyIter.next();
+countFileSize(kv.getValue());
+ }
+} catch (IOException ioEx) {
+ LOG.error("Unable to populate File Size Count in Recon DB. ", ioEx);
+ return new ImmutablePair<>(getTaskName(), false);
+}
+populateFileCountBySizeDB();
+
+LOG.info("Completed a 'reprocess' run of FileSizeCountTask.");
+return new ImmutablePair<>(getTaskName(), true);
+ }
+
+ @Override
+ protected Collection getTaskTables() {
+return tables;
+ }
+
+ void updateCountFromDB() {
+// Read - Write operations to DB are in ascending order
+// of file size upper bounds.
+List resultSet = fileCountBySizeDao.findAll();
+int index = 0;
+if (resultSet != null) {
+ for (FileCountBySize row : resultSet) {
+upperBoundCount[index] = row.getCount();
+index++;
+ }
+}
+ }
+
+ /**
+ * Read the Keys from update events and update the count of files
+ * pertaining to a certain upper bound.
+ *
+ * @param events Update events - PUT/DELETE.
+ * @return Pair
+ */
+ @Override
+ Pair process(OMUpdateEventBatch events) {
+
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r311792973
##
File path:
hadoop-ozone/ozone-recon-codegen/src/main/java/org/hadoop/ozone/recon/schema/UtilizationSchemaDefinition.java
##
@@ -65,5 +69,12 @@ void createClusterGrowthTable(Connection conn) {
.execute();
}
-
+ void createFileSizeCount(Connection conn) {
+DSL.using(conn).createTableIfNotExists(FILE_COUNT_BY_SIZE_TABLE_NAME)
+.column("file_size_kb", SQLDataType.BIGINT)
Review comment:
Sure.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
-
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r311792637
##
File path:
hadoop-ozone/ozone-recon/src/main/java/org/apache/hadoop/ozone/recon/tasks/FileSizeCountTask.java
##
@@ -0,0 +1,241 @@
+/**
+ * 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 org.apache.hadoop.ozone.recon.tasks;
+
+import com.google.inject.Inject;
+import org.apache.commons.lang3.tuple.ImmutablePair;
+import org.apache.commons.lang3.tuple.Pair;
+import org.apache.hadoop.ozone.om.OMMetadataManager;
+import org.apache.hadoop.ozone.om.helpers.OmKeyInfo;
+import org.apache.hadoop.utils.db.Table;
+import org.apache.hadoop.utils.db.TableIterator;
+import org.hadoop.ozone.recon.schema.tables.daos.FileCountBySizeDao;
+import org.hadoop.ozone.recon.schema.tables.pojos.FileCountBySize;
+import org.jooq.Configuration;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.Iterator;
+import java.util.List;
+
+/**
+ * Class to iterate over the OM DB and store the counts of existing/new
+ * files binned into ranges (1KB, 2Kb..,4MB,.., 1TB,..1PB) to the Recon
+ * fileSize DB.
+ */
+public class FileSizeCountTask extends ReconDBUpdateTask {
+ private static final Logger LOG =
+ LoggerFactory.getLogger(FileSizeCountTask.class);
+
+ private int maxBinSize = -1;
+ private long maxFileSizeUpperBound = 1125899906842624L; // 1 PB
+ private long[] upperBoundCount;
+ private long oneKb = 1024L;
+ private Collection tables = new ArrayList<>();
+ private FileCountBySizeDao fileCountBySizeDao;
+
+ @Inject
+ public FileSizeCountTask(OMMetadataManager omMetadataManager,
+ Configuration sqlConfiguration) {
+super("FileSizeCountTask");
+try {
+ tables.add(omMetadataManager.getKeyTable().getName());
+ fileCountBySizeDao = new FileCountBySizeDao(sqlConfiguration);
+} catch (Exception e) {
+ LOG.error("Unable to fetch Key Table updates ", e);
+}
+upperBoundCount = new long[getMaxBinSize()];
+ }
+
+ protected long getOneKB() {
+return oneKb;
+ }
+
+ protected long getMaxFileSizeUpperBound() {
+return maxFileSizeUpperBound;
+ }
+
+ protected int getMaxBinSize() {
+if (maxBinSize == -1) {
+ // extra bin to add files > 1PB.
+ maxBinSize = calculateBinIndex(maxFileSizeUpperBound) + 1;
+}
+return maxBinSize;
+ }
+
+ /**
+ * Read the Keys from OM snapshot DB and calculate the upper bound of
+ * File Size it belongs to.
+ *
+ * @param omMetadataManager OM Metadata instance.
+ * @return Pair
+ */
+ @Override
+ public Pair reprocess(OMMetadataManager omMetadataManager) {
+LOG.info("Starting a 'reprocess' run of FileSizeCountTask.");
+Table omKeyInfoTable = omMetadataManager.getKeyTable();
+try (TableIterator>
+keyIter = omKeyInfoTable.iterator()) {
+ while (keyIter.hasNext()) {
+Table.KeyValue kv = keyIter.next();
+countFileSize(kv.getValue());
+ }
+} catch (IOException ioEx) {
+ LOG.error("Unable to populate File Size Count in Recon DB. ", ioEx);
+ return new ImmutablePair<>(getTaskName(), false);
+}
+populateFileCountBySizeDB();
+
+LOG.info("Completed a 'reprocess' run of FileSizeCountTask.");
+return new ImmutablePair<>(getTaskName(), true);
+ }
+
+ @Override
+ protected Collection getTaskTables() {
+return tables;
+ }
+
+ void updateCountFromDB() {
+// Read - Write operations to DB are in ascending order
+// of file size upper bounds.
+List resultSet = fileCountBySizeDao.findAll();
+int index = 0;
+if (resultSet != null) {
+ for (FileCountBySize row : resultSet) {
+upperBoundCount[index] = row.getCount();
+index++;
+ }
+}
+ }
+
+ /**
+ * Read the Keys from update events and update the count of files
+ * pertaining to a certain upper bound.
+ *
+ * @param events Update events - PUT/DELETE.
+ * @return Pair
+ */
+ @Override
+ Pair process(OMUpdateEventBatch events) {
+
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r311792479
##
File path:
hadoop-ozone/ozone-recon/src/test/java/org/apache/hadoop/ozone/recon/api/TestUtilizationService.java
##
@@ -0,0 +1,108 @@
+/**
+ * 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 org.apache.hadoop.ozone.recon.api;
+
+import org.apache.hadoop.ozone.recon.ReconUtils;
+import org.hadoop.ozone.recon.schema.tables.daos.FileCountBySizeDao;
+import org.hadoop.ozone.recon.schema.tables.pojos.FileCountBySize;
+import org.junit.Test;
+import org.junit.runner.RunWith;
+import org.mockito.Mock;
+import org.powermock.core.classloader.annotations.PowerMockIgnore;
+import org.powermock.core.classloader.annotations.PrepareForTest;
+import org.powermock.modules.junit4.PowerMockRunner;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.List;
+
+import static org.junit.Assert.assertEquals;
+import static org.powermock.api.mockito.PowerMockito.mock;
+import static org.powermock.api.mockito.PowerMockito.when;
+import static org.mockito.Mockito.times;
+import static org.mockito.Mockito.verify;
+
+/**
+ * Test for File size count service.
+ */
+@RunWith(PowerMockRunner.class)
+@PowerMockIgnore({"javax.management.*", "javax.net.ssl.*"})
+@PrepareForTest(ReconUtils.class)
+public class TestUtilizationService {
+ private UtilizationService utilizationService;
+ @Mock private FileCountBySizeDao fileCountBySizeDao;
+ private List resultList = new ArrayList<>();
+ private int oneKb = 1024;
+ private int maxBinSize = 41;
+
+ public void setUpResultList() {
+for(int i = 0; i < 41; i++){
+ resultList.add(new FileCountBySize((long) Math.pow(2, (10+i)), (long)
i));
+}
+ }
+
+ @Test
+ public void testGetFileCounts() throws IOException {
+setUpResultList();
+
+utilizationService = mock(UtilizationService.class);
+when(utilizationService.getFileCounts()).thenCallRealMethod();
+when(utilizationService.getDao()).thenReturn(fileCountBySizeDao);
+when(fileCountBySizeDao.findAll()).thenReturn(resultList);
+
+utilizationService.getFileCounts();
+verify(utilizationService, times(1)).getFileCounts();
+verify(fileCountBySizeDao, times(1)).findAll();
+
+assertEquals(41, resultList.size());
+long fileSize = 4096L;
+int index = findIndex(fileSize);
+long count = resultList.get(index).getCount();
+assertEquals(index, count);
+
+fileSize = 1125899906842624L;
+index = findIndex(fileSize);
+if (index == Integer.MIN_VALUE) {
+ throw new IOException("File Size larger than permissible file size");
+}
+
+fileSize = 1025L;
+index = findIndex(fileSize);
+count = resultList.get(index).getCount();
+assertEquals(index, count);
+
+fileSize = 25L;
+index = findIndex(fileSize);
+count = resultList.get(index).getCount();
+assertEquals(index, count);
+ }
+
+ public int findIndex(long dataSize) {
+int logValue = (int) Math.ceil(Math.log(dataSize)/Math.log(2));
+if (logValue < 10) {
+ return 0;
+} else {
+ int index = logValue - 10;
+ if (index > maxBinSize) {
+return Integer.MIN_VALUE;
Review comment:
Sure.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
-
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r311341219
##
File path:
hadoop-ozone/ozone-recon/src/main/java/org/apache/hadoop/ozone/recon/tasks/FileSizeCountTask.java
##
@@ -155,70 +164,70 @@ private void fetchUpperBoundCount(String type) {
LOG.error("Unexpected exception while updating key data : {} {}",
updatedKey, e.getMessage());
return new ImmutablePair<>(getTaskName(), false);
- } finally {
-populateFileCountBySizeDB();
}
+ populateFileCountBySizeDB();
}
LOG.info("Completed a 'process' run of FileSizeCountTask.");
return new ImmutablePair<>(getTaskName(), true);
}
/**
* Calculate the bin index based on size of the Key.
+ * index is calculated as the number of right shifts
+ * needed until dataSize becomes zero.
*
* @param dataSize Size of the key.
* @return int bin index in upperBoundCount
*/
- private int calcBinIndex(long dataSize) {
-if(dataSize >= maxFileSizeUpperBound) {
- return Integer.MIN_VALUE;
-} else if (dataSize > SIZE_512_TB) {
- //given the small difference in 512TB and 512TB + 1B, index for both
would
- //return same, to differentiate specific condition added.
- return maxBinSize - 1;
-}
-int logValue = (int) Math.ceil(Math.log(dataSize)/Math.log(2));
-if(logValue < 10){
- return 0;
-} else{
- return (dataSize % ONE_KB == 0) ? logValue - 10 + 1: logValue - 10;
+ int calculateBinIndex(long dataSize) {
+int index = 0;
+while(dataSize != 0) {
+ dataSize >>= 1;
+ index += 1;
}
+return index < 10 ? 0 : index - 10;
}
- private void countFileSize(OmKeyInfo omKeyInfo) throws IOException{
-int index = calcBinIndex(omKeyInfo.getDataSize());
-if(index == Integer.MIN_VALUE) {
- throw new IOException("File Size larger than permissible file size "
- + maxFileSizeUpperBound +" bytes");
+ void countFileSize(OmKeyInfo omKeyInfo) {
+int index;
+if (omKeyInfo.getDataSize() >= maxFileSizeUpperBound) {
+ index = maxBinSize - 1;
+} else {
+ index = calculateBinIndex(omKeyInfo.getDataSize());
}
upperBoundCount[index]++;
}
- private void populateFileCountBySizeDB() {
+ /**
+ * Populate DB with the counts of file sizes calculated
+ * using the dao.
+ *
+ */
+ void populateFileCountBySizeDB() {
for (int i = 0; i < upperBoundCount.length; i++) {
long fileSizeUpperBound = (long) Math.pow(2, (10 + i));
FileCountBySize fileCountRecord =
fileCountBySizeDao.findById(fileSizeUpperBound);
FileCountBySize newRecord = new
FileCountBySize(fileSizeUpperBound, upperBoundCount[i]);
- if(fileCountRecord == null){
+ if (fileCountRecord == null) {
Review comment:
Done.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
-
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r311308069
##
File path:
hadoop-ozone/ozone-recon/src/main/java/org/apache/hadoop/ozone/recon/tasks/FileSizeCountTask.java
##
@@ -155,70 +164,70 @@ private void fetchUpperBoundCount(String type) {
LOG.error("Unexpected exception while updating key data : {} {}",
updatedKey, e.getMessage());
return new ImmutablePair<>(getTaskName(), false);
- } finally {
-populateFileCountBySizeDB();
}
+ populateFileCountBySizeDB();
}
LOG.info("Completed a 'process' run of FileSizeCountTask.");
return new ImmutablePair<>(getTaskName(), true);
}
/**
* Calculate the bin index based on size of the Key.
+ * index is calculated as the number of right shifts
+ * needed until dataSize becomes zero.
*
* @param dataSize Size of the key.
* @return int bin index in upperBoundCount
*/
- private int calcBinIndex(long dataSize) {
-if(dataSize >= maxFileSizeUpperBound) {
- return Integer.MIN_VALUE;
-} else if (dataSize > SIZE_512_TB) {
- //given the small difference in 512TB and 512TB + 1B, index for both
would
- //return same, to differentiate specific condition added.
- return maxBinSize - 1;
-}
-int logValue = (int) Math.ceil(Math.log(dataSize)/Math.log(2));
-if(logValue < 10){
- return 0;
-} else{
- return (dataSize % ONE_KB == 0) ? logValue - 10 + 1: logValue - 10;
+ int calculateBinIndex(long dataSize) {
+int index = 0;
+while(dataSize != 0) {
+ dataSize >>= 1;
+ index += 1;
}
+return index < 10 ? 0 : index - 10;
}
- private void countFileSize(OmKeyInfo omKeyInfo) throws IOException{
-int index = calcBinIndex(omKeyInfo.getDataSize());
-if(index == Integer.MIN_VALUE) {
- throw new IOException("File Size larger than permissible file size "
- + maxFileSizeUpperBound +" bytes");
+ void countFileSize(OmKeyInfo omKeyInfo) {
+int index;
+if (omKeyInfo.getDataSize() >= maxFileSizeUpperBound) {
+ index = maxBinSize - 1;
+} else {
+ index = calculateBinIndex(omKeyInfo.getDataSize());
}
upperBoundCount[index]++;
}
- private void populateFileCountBySizeDB() {
+ /**
+ * Populate DB with the counts of file sizes calculated
+ * using the dao.
+ *
+ */
+ void populateFileCountBySizeDB() {
for (int i = 0; i < upperBoundCount.length; i++) {
long fileSizeUpperBound = (long) Math.pow(2, (10 + i));
FileCountBySize fileCountRecord =
fileCountBySizeDao.findById(fileSizeUpperBound);
FileCountBySize newRecord = new
FileCountBySize(fileSizeUpperBound, upperBoundCount[i]);
- if(fileCountRecord == null){
+ if (fileCountRecord == null) {
fileCountBySizeDao.insert(newRecord);
- } else{
+ } else {
fileCountBySizeDao.update(newRecord);
}
}
}
private void updateUpperBoundCount(OmKeyInfo value, String operation)
throws IOException {
-int binIndex = calcBinIndex(value.getDataSize());
-if(binIndex == Integer.MIN_VALUE) {
+int binIndex = calculateBinIndex(value.getDataSize());
+if (binIndex == Integer.MIN_VALUE) {
Review comment:
Yes, it was from a previous check where there was an exception for fileSize
> permitted value of 1 PB.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
-
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r311308884
##
File path:
hadoop-ozone/ozone-recon/src/main/java/org/apache/hadoop/ozone/recon/tasks/FileSizeCountTask.java
##
@@ -155,70 +164,70 @@ private void fetchUpperBoundCount(String type) {
LOG.error("Unexpected exception while updating key data : {} {}",
updatedKey, e.getMessage());
return new ImmutablePair<>(getTaskName(), false);
- } finally {
-populateFileCountBySizeDB();
}
+ populateFileCountBySizeDB();
}
LOG.info("Completed a 'process' run of FileSizeCountTask.");
return new ImmutablePair<>(getTaskName(), true);
}
/**
* Calculate the bin index based on size of the Key.
+ * index is calculated as the number of right shifts
+ * needed until dataSize becomes zero.
*
* @param dataSize Size of the key.
* @return int bin index in upperBoundCount
*/
- private int calcBinIndex(long dataSize) {
-if(dataSize >= maxFileSizeUpperBound) {
- return Integer.MIN_VALUE;
-} else if (dataSize > SIZE_512_TB) {
- //given the small difference in 512TB and 512TB + 1B, index for both
would
- //return same, to differentiate specific condition added.
- return maxBinSize - 1;
-}
-int logValue = (int) Math.ceil(Math.log(dataSize)/Math.log(2));
-if(logValue < 10){
- return 0;
-} else{
- return (dataSize % ONE_KB == 0) ? logValue - 10 + 1: logValue - 10;
+ int calculateBinIndex(long dataSize) {
+int index = 0;
+while(dataSize != 0) {
+ dataSize >>= 1;
+ index += 1;
}
+return index < 10 ? 0 : index - 10;
}
- private void countFileSize(OmKeyInfo omKeyInfo) throws IOException{
-int index = calcBinIndex(omKeyInfo.getDataSize());
-if(index == Integer.MIN_VALUE) {
- throw new IOException("File Size larger than permissible file size "
- + maxFileSizeUpperBound +" bytes");
+ void countFileSize(OmKeyInfo omKeyInfo) {
+int index;
+if (omKeyInfo.getDataSize() >= maxFileSizeUpperBound) {
+ index = maxBinSize - 1;
+} else {
+ index = calculateBinIndex(omKeyInfo.getDataSize());
}
upperBoundCount[index]++;
}
- private void populateFileCountBySizeDB() {
+ /**
+ * Populate DB with the counts of file sizes calculated
+ * using the dao.
+ *
+ */
+ void populateFileCountBySizeDB() {
for (int i = 0; i < upperBoundCount.length; i++) {
long fileSizeUpperBound = (long) Math.pow(2, (10 + i));
FileCountBySize fileCountRecord =
fileCountBySizeDao.findById(fileSizeUpperBound);
FileCountBySize newRecord = new
FileCountBySize(fileSizeUpperBound, upperBoundCount[i]);
- if(fileCountRecord == null){
+ if (fileCountRecord == null) {
Review comment:
Sure, it is an extra bin to add files > maxFileSizeUpperBound.
Also, did you mean LONG.MAX_VALUE?
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
-
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r311308096
##
File path:
hadoop-ozone/ozone-recon/src/main/java/org/apache/hadoop/ozone/recon/tasks/FileSizeCountTask.java
##
@@ -155,70 +164,70 @@ private void fetchUpperBoundCount(String type) {
LOG.error("Unexpected exception while updating key data : {} {}",
updatedKey, e.getMessage());
return new ImmutablePair<>(getTaskName(), false);
- } finally {
-populateFileCountBySizeDB();
}
+ populateFileCountBySizeDB();
}
LOG.info("Completed a 'process' run of FileSizeCountTask.");
return new ImmutablePair<>(getTaskName(), true);
}
/**
* Calculate the bin index based on size of the Key.
+ * index is calculated as the number of right shifts
+ * needed until dataSize becomes zero.
*
* @param dataSize Size of the key.
* @return int bin index in upperBoundCount
*/
- private int calcBinIndex(long dataSize) {
-if(dataSize >= maxFileSizeUpperBound) {
- return Integer.MIN_VALUE;
-} else if (dataSize > SIZE_512_TB) {
- //given the small difference in 512TB and 512TB + 1B, index for both
would
- //return same, to differentiate specific condition added.
- return maxBinSize - 1;
-}
-int logValue = (int) Math.ceil(Math.log(dataSize)/Math.log(2));
-if(logValue < 10){
- return 0;
-} else{
- return (dataSize % ONE_KB == 0) ? logValue - 10 + 1: logValue - 10;
+ int calculateBinIndex(long dataSize) {
+int index = 0;
+while(dataSize != 0) {
+ dataSize >>= 1;
+ index += 1;
}
+return index < 10 ? 0 : index - 10;
Review comment:
Sure.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
-
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r311308069
##
File path:
hadoop-ozone/ozone-recon/src/main/java/org/apache/hadoop/ozone/recon/tasks/FileSizeCountTask.java
##
@@ -155,70 +164,70 @@ private void fetchUpperBoundCount(String type) {
LOG.error("Unexpected exception while updating key data : {} {}",
updatedKey, e.getMessage());
return new ImmutablePair<>(getTaskName(), false);
- } finally {
-populateFileCountBySizeDB();
}
+ populateFileCountBySizeDB();
}
LOG.info("Completed a 'process' run of FileSizeCountTask.");
return new ImmutablePair<>(getTaskName(), true);
}
/**
* Calculate the bin index based on size of the Key.
+ * index is calculated as the number of right shifts
+ * needed until dataSize becomes zero.
*
* @param dataSize Size of the key.
* @return int bin index in upperBoundCount
*/
- private int calcBinIndex(long dataSize) {
-if(dataSize >= maxFileSizeUpperBound) {
- return Integer.MIN_VALUE;
-} else if (dataSize > SIZE_512_TB) {
- //given the small difference in 512TB and 512TB + 1B, index for both
would
- //return same, to differentiate specific condition added.
- return maxBinSize - 1;
-}
-int logValue = (int) Math.ceil(Math.log(dataSize)/Math.log(2));
-if(logValue < 10){
- return 0;
-} else{
- return (dataSize % ONE_KB == 0) ? logValue - 10 + 1: logValue - 10;
+ int calculateBinIndex(long dataSize) {
+int index = 0;
+while(dataSize != 0) {
+ dataSize >>= 1;
+ index += 1;
}
+return index < 10 ? 0 : index - 10;
}
- private void countFileSize(OmKeyInfo omKeyInfo) throws IOException{
-int index = calcBinIndex(omKeyInfo.getDataSize());
-if(index == Integer.MIN_VALUE) {
- throw new IOException("File Size larger than permissible file size "
- + maxFileSizeUpperBound +" bytes");
+ void countFileSize(OmKeyInfo omKeyInfo) {
+int index;
+if (omKeyInfo.getDataSize() >= maxFileSizeUpperBound) {
+ index = maxBinSize - 1;
+} else {
+ index = calculateBinIndex(omKeyInfo.getDataSize());
}
upperBoundCount[index]++;
}
- private void populateFileCountBySizeDB() {
+ /**
+ * Populate DB with the counts of file sizes calculated
+ * using the dao.
+ *
+ */
+ void populateFileCountBySizeDB() {
for (int i = 0; i < upperBoundCount.length; i++) {
long fileSizeUpperBound = (long) Math.pow(2, (10 + i));
FileCountBySize fileCountRecord =
fileCountBySizeDao.findById(fileSizeUpperBound);
FileCountBySize newRecord = new
FileCountBySize(fileSizeUpperBound, upperBoundCount[i]);
- if(fileCountRecord == null){
+ if (fileCountRecord == null) {
fileCountBySizeDao.insert(newRecord);
- } else{
+ } else {
fileCountBySizeDao.update(newRecord);
}
}
}
private void updateUpperBoundCount(OmKeyInfo value, String operation)
throws IOException {
-int binIndex = calcBinIndex(value.getDataSize());
-if(binIndex == Integer.MIN_VALUE) {
+int binIndex = calculateBinIndex(value.getDataSize());
+if (binIndex == Integer.MIN_VALUE) {
Review comment:
Yes, it was from a previous check where there was an exception for fileSize
> permitted value of 1 B.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
-
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r311286958
##
File path:
hadoop-ozone/ozone-recon/src/main/java/org/apache/hadoop/ozone/recon/tasks/FileSizeCountTask.java
##
@@ -0,0 +1,254 @@
+/**
+ * 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 org.apache.hadoop.ozone.recon.tasks;
+
+import com.google.inject.Inject;
+import org.apache.commons.lang3.tuple.ImmutablePair;
+import org.apache.commons.lang3.tuple.Pair;
+import org.apache.hadoop.ozone.om.OMMetadataManager;
+import org.apache.hadoop.ozone.om.helpers.OmKeyInfo;
+import org.apache.hadoop.utils.db.Table;
+import org.apache.hadoop.utils.db.TableIterator;
+import org.hadoop.ozone.recon.schema.tables.daos.FileCountBySizeDao;
+import org.hadoop.ozone.recon.schema.tables.pojos.FileCountBySize;
+import org.jooq.Configuration;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.Iterator;
+import java.util.List;
+
+/**
+ * Class to iterate over the OM DB and store the counts of existing/new
+ * files binned into ranges (1KB, 10Kb..,10MB,..1PB) to the Recon
+ * fileSize DB.
+ */
+public class FileSizeCountTask extends ReconDBUpdateTask {
+ private static final Logger LOG =
+ LoggerFactory.getLogger(FileSizeCountTask.class);
+
+ private int maxBinSize;
+ private long maxFileSizeUpperBound = 1125899906842624L; // 1 PB
+ private long[] upperBoundCount = new long[maxBinSize];
+ private long ONE_KB = 1024L;
+ private Collection tables = new ArrayList<>();
+ private FileCountBySizeDao fileCountBySizeDao;
+
+ @Inject
+ public FileSizeCountTask(OMMetadataManager omMetadataManager,
+ Configuration sqlConfiguration) {
+super("FileSizeCountTask");
+try {
+ tables.add(omMetadataManager.getKeyTable().getName());
+ fileCountBySizeDao = new FileCountBySizeDao(sqlConfiguration);
+} catch (Exception e) {
+ LOG.error("Unable to fetch Key Table updates ", e);
+}
+ }
+
+ protected long getOneKB() {
+return ONE_KB;
+ }
+
+ protected long getMaxFileSizeUpperBound() {
+return maxFileSizeUpperBound;
+ }
+
+ protected int getMaxBinSize() {
+return maxBinSize;
+ }
+
+ /**
+ * Read the Keys from OM snapshot DB and calculate the upper bound of
+ * File Size it belongs to.
+ *
+ * @param omMetadataManager OM Metadata instance.
+ * @return Pair
+ */
+ @Override
+ public Pair reprocess(OMMetadataManager omMetadataManager) {
+LOG.info("Starting a 'reprocess' run of FileSizeCountTask.");
+
+fetchUpperBoundCount("reprocess");
+
+Table omKeyInfoTable = omMetadataManager.getKeyTable();
+try (TableIterator>
+keyIter = omKeyInfoTable.iterator()) {
+ while (keyIter.hasNext()) {
+Table.KeyValue kv = keyIter.next();
+countFileSize(kv.getValue());
+ }
+} catch (IOException ioEx) {
+ LOG.error("Unable to populate File Size Count in Recon DB. ", ioEx);
+ return new ImmutablePair<>(getTaskName(), false);
+} finally {
+ populateFileCountBySizeDB();
+}
+
+LOG.info("Completed a 'reprocess' run of FileSizeCountTask.");
+return new ImmutablePair<>(getTaskName(), true);
+ }
+
+ void setMaxBinSize() {
+maxBinSize = (int)(long) (Math.log(getMaxFileSizeUpperBound())
+/Math.log(2)) - 10;
+maxBinSize += 2; // extra bin to add files > 1PB.
+ }
+
+ void fetchUpperBoundCount(String type) {
+setMaxBinSize();
+if (type.equals("process")) {
+ //update array with file size count from DB
+ List resultSet = fileCountBySizeDao.findAll();
+ int index = 0;
+ if (resultSet != null) {
+for (FileCountBySize row : resultSet) {
+ upperBoundCount[index] = row.getCount();
+ index++;
+}
+ }
+} else {
+ upperBoundCount = new long[getMaxBinSize()];//initialize array
+}
+ }
+
+ @Override
+ protected Collection getTaskTables() {
+return tables;
+ }
+
+ /**
+ * Read the Keys from update events and
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r310805057
##
File path:
hadoop-ozone/ozone-recon/src/main/java/org/apache/hadoop/ozone/recon/tasks/FileSizeCountTask.java
##
@@ -0,0 +1,254 @@
+/**
+ * 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 org.apache.hadoop.ozone.recon.tasks;
+
+import com.google.inject.Inject;
+import org.apache.commons.lang3.tuple.ImmutablePair;
+import org.apache.commons.lang3.tuple.Pair;
+import org.apache.hadoop.ozone.om.OMMetadataManager;
+import org.apache.hadoop.ozone.om.helpers.OmKeyInfo;
+import org.apache.hadoop.utils.db.Table;
+import org.apache.hadoop.utils.db.TableIterator;
+import org.hadoop.ozone.recon.schema.tables.daos.FileCountBySizeDao;
+import org.hadoop.ozone.recon.schema.tables.pojos.FileCountBySize;
+import org.jooq.Configuration;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.Iterator;
+import java.util.List;
+
+/**
+ * Class to iterate over the OM DB and store the counts of existing/new
+ * files binned into ranges (1KB, 10Kb..,10MB,..1PB) to the Recon
+ * fileSize DB.
+ */
+public class FileSizeCountTask extends ReconDBUpdateTask {
+ private static final Logger LOG =
+ LoggerFactory.getLogger(FileSizeCountTask.class);
+
+ private int maxBinSize;
+ private long maxFileSizeUpperBound = 1125899906842624L; // 1 PB
+ private long[] upperBoundCount = new long[maxBinSize];
+ private long ONE_KB = 1024L;
+ private Collection tables = new ArrayList<>();
+ private FileCountBySizeDao fileCountBySizeDao;
+
+ @Inject
+ public FileSizeCountTask(OMMetadataManager omMetadataManager,
+ Configuration sqlConfiguration) {
+super("FileSizeCountTask");
+try {
+ tables.add(omMetadataManager.getKeyTable().getName());
+ fileCountBySizeDao = new FileCountBySizeDao(sqlConfiguration);
+} catch (Exception e) {
+ LOG.error("Unable to fetch Key Table updates ", e);
+}
+ }
+
+ protected long getOneKB() {
+return ONE_KB;
+ }
+
+ protected long getMaxFileSizeUpperBound() {
+return maxFileSizeUpperBound;
+ }
+
+ protected int getMaxBinSize() {
+return maxBinSize;
+ }
+
+ /**
+ * Read the Keys from OM snapshot DB and calculate the upper bound of
+ * File Size it belongs to.
+ *
+ * @param omMetadataManager OM Metadata instance.
+ * @return Pair
+ */
+ @Override
+ public Pair reprocess(OMMetadataManager omMetadataManager) {
+LOG.info("Starting a 'reprocess' run of FileSizeCountTask.");
+
+fetchUpperBoundCount("reprocess");
+
+Table omKeyInfoTable = omMetadataManager.getKeyTable();
+try (TableIterator>
+keyIter = omKeyInfoTable.iterator()) {
+ while (keyIter.hasNext()) {
+Table.KeyValue kv = keyIter.next();
+countFileSize(kv.getValue());
+ }
+} catch (IOException ioEx) {
+ LOG.error("Unable to populate File Size Count in Recon DB. ", ioEx);
+ return new ImmutablePair<>(getTaskName(), false);
+} finally {
+ populateFileCountBySizeDB();
+}
+
+LOG.info("Completed a 'reprocess' run of FileSizeCountTask.");
+return new ImmutablePair<>(getTaskName(), true);
+ }
+
+ void setMaxBinSize() {
+maxBinSize = (int)(long) (Math.log(getMaxFileSizeUpperBound())
+/Math.log(2)) - 10;
+maxBinSize += 2; // extra bin to add files > 1PB.
+ }
+
+ void fetchUpperBoundCount(String type) {
+setMaxBinSize();
+if (type.equals("process")) {
+ //update array with file size count from DB
+ List resultSet = fileCountBySizeDao.findAll();
+ int index = 0;
+ if (resultSet != null) {
+for (FileCountBySize row : resultSet) {
+ upperBoundCount[index] = row.getCount();
+ index++;
+}
+ }
+} else {
+ upperBoundCount = new long[getMaxBinSize()];//initialize array
+}
+ }
+
+ @Override
+ protected Collection getTaskTables() {
+return tables;
+ }
+
+ /**
+ * Read the Keys from update events and
[GitHub] [hadoop] shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add ability in Recon to track the number of small files in an Ozone Cluster
shwetayakkali commented on a change in pull request #1146: HDDS-1366. Add
ability in Recon to track the number of small files in an Ozone Cluster
URL: https://github.com/apache/hadoop/pull/1146#discussion_r307603546
##
File path:
hadoop-ozone/ozone-recon/src/test/java/org/apache/hadoop/ozone/recon/tasks/TestFileSizeCountTask.java
##
@@ -0,0 +1,311 @@
+/**
+ * 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 org.apache.hadoop.ozone.recon.tasks;
+
+import com.google.inject.Injector;
+import org.apache.hadoop.hdds.client.BlockID;
+import org.apache.hadoop.hdds.conf.OzoneConfiguration;
+import org.apache.hadoop.hdds.protocol.proto.HddsProtos;
+import org.apache.hadoop.hdds.scm.pipeline.Pipeline;
+import org.apache.hadoop.ozone.om.OMMetadataManager;
+import org.apache.hadoop.ozone.om.helpers.OmKeyInfo;
+import org.apache.hadoop.ozone.om.helpers.OmKeyLocationInfo;
+import org.apache.hadoop.ozone.om.helpers.OmKeyLocationInfoGroup;
+import org.apache.hadoop.ozone.recon.AbstractOMMetadataManagerTest;
+import org.apache.hadoop.ozone.recon.GuiceInjectorUtilsForTestsImpl;
+import org.apache.hadoop.ozone.recon.ReconUtils;
+import org.apache.hadoop.ozone.recon.recovery.ReconOMMetadataManager;
+import org.apache.hadoop.ozone.recon.spi.impl.OzoneManagerServiceProviderImpl;
+import org.apache.hadoop.utils.db.Table;
+import org.hadoop.ozone.recon.schema.UtilizationSchemaDefinition;
+import org.hadoop.ozone.recon.schema.tables.daos.FileCountBySizeDao;
+import org.hadoop.ozone.recon.schema.tables.pojos.FileCountBySize;
+import org.jooq.Configuration;
+import org.jooq.impl.DSL;
+import org.jooq.impl.DefaultConfiguration;
+import org.junit.Before;
+import org.junit.Rule;
+import org.junit.Test;
+import org.junit.rules.TemporaryFolder;
+import org.junit.runner.RunWith;
+import org.powermock.core.classloader.annotations.PowerMockIgnore;
+import org.powermock.core.classloader.annotations.PrepareForTest;
+import org.powermock.modules.junit4.PowerMockRunner;
+
+import javax.sql.DataSource;
+import java.io.IOException;
+import java.util.*;
+
+import static org.junit.Assert.*;
+
+/**
+ * Unit test for Container Key mapper task.
+ */
+@RunWith(PowerMockRunner.class)
+@PowerMockIgnore({"javax.management.*", "javax.net.ssl.*"})
+@PrepareForTest(ReconUtils.class)
Review comment:
Nothing yet
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
-
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
