/*
* 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.chukwa.extraction.demux;
import java.io.IOException;
import java.net.URI;
import java.text.SimpleDateFormat;
import java.util.Calendar;
import org.apache.hadoop.chukwa.conf.ChukwaConfiguration;
import org.apache.hadoop.fs.FileStatus;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.FileUtil;
import org.apache.hadoop.fs.Path;
import org.apache.log4j.Logger;
// TODO
// First version of the Spill
// need some polishing
public class MoveToRepository {
static Logger log = Logger.getLogger(MoveToRepository.class);
static ChukwaConfiguration conf = null;
static FileSystem fs = null;
static final String HadoopLogDir = "_logs";
static final String hadoopTempDir = "_temporary";
static SimpleDateFormat sdf = new java.text.SimpleDateFormat("yyyyMMdd");
static Calendar calendar = Calendar.getInstance();
static void processClutserDirectory(Path srcDir, String destDir)
throws Exception {
log.info("processClutserDirectory (" + srcDir.getName() + "," + destDir
+ ")");
FileStatus fstat = fs.getFileStatus(srcDir);
if (!fstat.isDir()) {
throw new IOException(srcDir + " is not a directory!");
} else {
FileStatus[] datasourceDirectories = fs.listStatus(srcDir);
for (FileStatus datasourceDirectory : datasourceDirectories) {
log.info(datasourceDirectory.getPath() + " isDir?"
+ datasourceDirectory.isDir());
if (!datasourceDirectory.isDir()) {
throw new IOException(
"Top level datasource directory should be a directory :"
+ datasourceDirectory.getPath());
}
String dirName = datasourceDirectory.getPath().getName();
Path destPath = new Path(destDir + "/" + dirName);
log.info("dest directory path: " + destPath);
log.info("processClutserDirectory processing Datasource: (" + dirName
+ ")");
processDatasourceDirectory(srcDir.getName(), datasourceDirectory
.getPath(), destDir + "/" + dirName);
}
}
}
static void processDatasourceDirectory(String cluster, Path srcDir,
String destDir) throws Exception {
String fileName = null;
int fileDay = 0;
int fileHour = 0;
int fileMin = 0;
FileStatus[] recordFiles = fs.listStatus(srcDir);
for (FileStatus recordFile : recordFiles) {
// dataSource_20080915_18_15.1.evt
// <datasource>_<yyyyMMdd_HH_mm>.1.evt
fileName = recordFile.getPath().getName();
log.info("processDatasourceDirectory processing RecordFile: (" + fileName
+ ")");
log.info("fileName: " + fileName);
int l = fileName.length();
String dataSource = srcDir.getName();
log.info("Datasource: " + dataSource);
if (fileName.endsWith(".D.evt")) {
// Hadoop_dfs_datanode_20080919.D.evt
fileDay = Integer.parseInt(fileName.substring(l - 14, l - 6));
writeRecordFile(destDir + "/" + fileDay + "/", recordFile.getPath(),
dataSource + "_" + fileDay);
// mark this directory for Daily rotate (re-process)
addDirectory4Rolling(true, fileDay, fileHour, cluster, dataSource);
} else if (fileName.endsWith(".H.evt")) {
// Hadoop_dfs_datanode_20080925_1.H.evt
// Hadoop_dfs_datanode_20080925_12.H.evt
String day = null;
String hour = null;
if (fileName.charAt(l - 8) == '_') {
day = fileName.substring(l - 16, l - 8);
log.info("day->" + day);
hour = "" + fileName.charAt(l - 7);
log.info("hour->" + hour);
} else {
day = fileName.substring(l - 17, l - 9);
log.info("day->" + day);
hour = fileName.substring(l - 8, l - 6);
log.info("hour->" + hour);
}
fileDay = Integer.parseInt(day);
fileHour = Integer.parseInt(hour);
// rotate there so spill
writeRecordFile(destDir + "/" + fileDay + "/" + fileHour + "/",
recordFile.getPath(), dataSource + "_" + fileDay + "_" + fileHour);
// mark this directory for daily rotate
addDirectory4Rolling(true, fileDay, fileHour, cluster, dataSource);
} else if (fileName.endsWith(".R.evt")) {
if (fileName.charAt(l - 11) == '_') {
fileDay = Integer.parseInt(fileName.substring(l - 19, l - 11));
fileHour = Integer.parseInt("" + fileName.charAt(l - 10));
fileMin = Integer.parseInt(fileName.substring(l - 8, l - 6));
} else {
fileDay = Integer.parseInt(fileName.substring(l - 20, l - 12));
fileHour = Integer.parseInt(fileName.substring(l - 11, l - 9));
fileMin = Integer.parseInt(fileName.substring(l - 8, l - 6));
}
log.info("fileDay: " + fileDay);
log.info("fileHour: " + fileHour);
log.info("fileMin: " + fileMin);
writeRecordFile(destDir + "/" + fileDay + "/" + fileHour + "/"
+ fileMin, recordFile.getPath(), dataSource + "_" + fileDay + "_"
+ fileHour + "_" + fileMin);
// mark this directory for hourly rotate
addDirectory4Rolling(false, fileDay, fileHour, cluster, dataSource);
} else {
throw new RuntimeException("Wrong fileName format! [" + fileName + "]");
}
}
}
static void addDirectory4Rolling(boolean isDailyOnly, int day, int hour,
String cluster, String dataSource) throws IOException {
// TODO get root directory from config
String rollingDirectory = "/chukwa/rolling/";
Path path = new Path(rollingDirectory + "/daily/" + day + "/" + cluster
+ "/" + dataSource);
if (!fs.exists(path)) {
fs.mkdirs(path);
}
if (!isDailyOnly) {
path = new Path(rollingDirectory + "/hourly/" + day + "/" + hour + "/"
+ cluster + "/" + dataSource);
if (!fs.exists(path)) {
fs.mkdirs(path);
}
}
}
static void writeRecordFile(String destDir, Path recordFile, String fileName)
throws IOException {
boolean done = false;
int count = 1;
do {
Path destDirPath = new Path(destDir);
Path destFilePath = new Path(destDir + "/" + fileName + "." + count
+ ".evt");
if (!fs.exists(destDirPath)) {
fs.mkdirs(destDirPath);
log.info(">>>>>>>>>>>> create Dir" + destDirPath);
}
if (!fs.exists(destFilePath)) {
log.info(">>>>>>>>>>>> Before Rename" + recordFile + " -- "
+ destFilePath);
boolean rename = fs.rename(recordFile,destFilePath);
done = true;
log.info(">>>>>>>>>>>> after Rename" + destFilePath + " , rename:"+rename);
} else {
log.info("increment spill for :" + destFilePath);
}
count++;
if (count > 1000) {
log.warn("too many files in this directory: " + destDir);
}
} while (!done);
}
static boolean checkRotate(String directoryAsString,
boolean createDirectoryIfNotExist) throws IOException {
Path directory = new Path(directoryAsString);
boolean exist = fs.exists(directory);
if (!exist) {
if (createDirectoryIfNotExist == true) {
fs.mkdirs(directory);
}
return false;
} else {
return fs.exists(new Path(directoryAsString + "/rotateDone"));
}
}
/**
* @param args
* @throws Exception
*/
public static void main(String[] args) throws Exception {
conf = new ChukwaConfiguration();
String fsName = conf.get("writer.hdfs.filesystem");
fs = FileSystem.get(new URI(fsName), conf);
Path srcDir = new Path(args[0]);
String destDir = args[1];
log.info("Start MoveToRepository main()");
FileStatus fstat = fs.getFileStatus(srcDir);
if (!fstat.isDir()) {
throw new IOException(srcDir + " is not a directory!");
} else {
FileStatus[] clusters = fs.listStatus(srcDir);
// Run a moveOrMerge on all clusters
String name = null;
for (FileStatus cluster : clusters) {
name = cluster.getPath().getName();
// Skip hadoop M/R outputDir
if ((name.intern() == HadoopLogDir.intern())
|| (name.intern() == hadoopTempDir.intern())) {
continue;
}
log
.info("main procesing Cluster (" + cluster.getPath().getName()
+ ")");
processClutserDirectory(cluster.getPath(), destDir + "/"
+ cluster.getPath().getName());
// Delete the demux's cluster dir
FileUtil.fullyDelete(fs, cluster.getPath());
}
}
log.info("Done with MoveToRepository main()");
}
}