ActiveFile(File f, List<Thread> list) {
file = f;
if (list != null) {
threads.addAll(list);
}
threads.add(Thread.currentThread());
}
void checkFormat(Job job) throws Exception {
TaskAttemptContext attemptContext =
new TaskAttemptContext(job.getConfiguration(), new TaskAttemptID("123", 0, false, 1, 2));
MyClassMessagePackBase64LineInputFormat format = new MyClassMessagePackBase64LineInputFormat();
FileInputFormat.setInputPaths(job, workDir);
List<InputSplit> splits = format.getSplits(job);
for (int j = 0; j < splits.size(); j++) {
RecordReader<LongWritable, MyClassWritable> reader =
format.createRecordReader(splits.get(j), attemptContext);
reader.initialize(splits.get(j), attemptContext);
int count = 0;
try {
while (reader.nextKeyValue()) {
LongWritable key = reader.getCurrentKey();
MyClassWritable val = reader.getCurrentValue();
MyClass mc = val.get();
assertEquals(mc.v, count);
assertEquals(mc.s, Integer.toString(count));
count++;
}
} finally {
reader.close();
}
}
}
protected static double Distance(List<Double> p1, List<Double> p2) {
double sumOfSquaredDifferences = 0.0;
for (int i = 0; i < p1.size(); i++) {
sumOfSquaredDifferences += Math.pow(p1.get(i) - p2.get(i), 2.0);
}
return Math.pow(sumOfSquaredDifferences, 0.5);
}
public static List<Double> GetPoint(String s) {
StringTokenizer tokenizer = new StringTokenizer(s);
List<Double> p = new ArrayList<Double>(58);
while (tokenizer.hasMoreTokens()) {
p.add(Double.parseDouble(tokenizer.nextToken()));
}
return p;
}
public static List<String> tokenizeString(Analyzer analyzer, String string) {
List<String> result = new ArrayList<String>();
try {
TokenStream stream = analyzer.tokenStream(null, new StringReader(string));
stream.reset();
while (stream.incrementToken()) {
result.add(stream.getAttribute(CharTermAttribute.class).toString());
}
} catch (IOException e) {
// not thrown b/c we're using a string reader...
throw new RuntimeException(e);
}
return result;
}
public static List<InetSocketAddress> readAddresses(Path path, Configuration conf)
throws IOException {
final List<InetSocketAddress> addrs = new ArrayList<InetSocketAddress>();
for (final String line : readConfig(path, conf)) {
final StringTokenizer tokens = new StringTokenizer(line);
if (tokens.hasMoreTokens()) {
final String host = tokens.nextToken();
if (tokens.hasMoreTokens()) {
final String port = tokens.nextToken();
addrs.add(new InetSocketAddress(host, Integer.parseInt(port)));
}
}
}
return addrs;
}
public void configure(JobConf conf) {
numberOfCenters = Integer.valueOf(conf.get("numberOfCenters"));
centersDirectory = conf.get("centersReadDirectory");
try {
Configuration c = new Configuration();
FileSystem fs = FileSystem.get(c);
for (int index = 0; index < numberOfCenters; ++index) {
SequenceFile.Reader reader =
new SequenceFile.Reader(fs, new Path(centersDirectory + "/centers/" + index), c);
LongWritable key = new LongWritable();
Point value = new Point();
reader.next(key, value);
Point center = (Point) value;
centers.add(center);
reader.close();
}
} catch (IOException e) {
// do nothing
// I hope this doesn't happen
System.out.println("well, damn.");
e.printStackTrace();
}
}
/**
* Generate the list of files and make them into FileSplits. This needs to be copied to insert a
* filter on acceptable data
*/
@Override
public List<InputSplit> getSplits(JobContext job) throws IOException {
long minSize = Math.max(getFormatMinSplitSize(), getMinSplitSize(job));
long maxSize = getMaxSplitSize(job);
long desiredMappers =
job.getConfiguration().getLong("org.systemsbiology.jxtandem.DesiredXMLInputMappers", 0);
// generate splits
List<InputSplit> splits = new ArrayList<InputSplit>();
List<FileStatus> fileStatuses = listStatus(job);
boolean forceNumberMappers = fileStatuses.size() == 1;
for (FileStatus file : fileStatuses) {
Path path = file.getPath();
if (!isPathAcceptable(path)) // filter acceptable data
continue;
FileSystem fs = path.getFileSystem(job.getConfiguration());
long length = file.getLen();
BlockLocation[] blkLocations = fs.getFileBlockLocations(file, 0, length);
if ((length != 0) && isSplitable(job, path)) {
long blockSize = file.getBlockSize();
// use desired mappers to force more splits
if (forceNumberMappers && desiredMappers > 0)
maxSize = Math.min(maxSize, (length / desiredMappers));
long splitSize = computeSplitSize(blockSize, minSize, maxSize);
long bytesRemaining = length;
while (withinSlop(splitSize, bytesRemaining)) {
int blkIndex = getBlockIndex(blkLocations, length - bytesRemaining);
splits.add(
new FileSplit(
path, length - bytesRemaining, splitSize, blkLocations[blkIndex].getHosts()));
bytesRemaining -= splitSize;
}
if (bytesRemaining != 0) {
splits.add(
new FileSplit(
path,
length - bytesRemaining,
bytesRemaining,
blkLocations[blkLocations.length - 1].getHosts()));
}
} else if (length != 0) {
splits.add(new FileSplit(path, 0, length, blkLocations[0].getHosts()));
} else {
// Create empty hosts array for zero length files
splits.add(new FileSplit(path, 0, length, new String[0]));
}
}
System.out.println("Total # of splits: " + splits.size());
// LOG.debug("Total # of splits: " + splits.size());
return splits;
}
public List<String> resolve(List<String> names) {
List<String> m = new ArrayList<String>(names.size());
if (names.isEmpty()) {
return m;
}
if (scriptName == null) {
for (int i = 0; i < names.size(); i++) {
m.add(NetworkTopology.DEFAULT_RACK);
}
return m;
}
String output = runResolveCommand(names);
if (output != null) {
StringTokenizer allSwitchInfo = new StringTokenizer(output);
while (allSwitchInfo.hasMoreTokens()) {
String switchInfo = allSwitchInfo.nextToken();
m.add(switchInfo);
}
if (m.size() != names.size()) {
// invalid number of entries returned by the script
LOG.warn(
"Script "
+ scriptName
+ " returned "
+ Integer.toString(m.size())
+ " values when "
+ Integer.toString(names.size())
+ " were expected.");
return null;
}
} else {
// an error occurred. return null to signify this.
// (exn was already logged in runResolveCommand)
return null;
}
return m;
}
public void map(
LongWritable key, Point value, OutputCollector<LongWritable, Point> output, Reporter reporter)
throws IOException {
double min = value.sumOfSquares(centers.get(0));
int best = 0;
for (int index = 1; index < numberOfCenters; ++index) {
double current = value.sumOfSquares(centers.get(index));
if (current < min) {
min = current;
best = index;
}
}
reporter.incrCounter("NUMBER", "NODES", 1);
reporter.incrCounter("CENTER", "" + best, 1);
output.collect(new LongWritable(best), value);
}
private static void deleteCache(Configuration conf, MRAsyncDiskService asyncDiskService)
throws IOException {
List<CacheStatus> deleteSet = new LinkedList<CacheStatus>();
// try deleting cache Status with refcount of zero
synchronized (cachedArchives) {
for (Iterator<String> it = cachedArchives.keySet().iterator(); it.hasNext(); ) {
String cacheId = (String) it.next();
CacheStatus lcacheStatus = cachedArchives.get(cacheId);
if (lcacheStatus.refcount == 0) {
// delete this cache entry from the global list
// and mark the localized file for deletion
deleteSet.add(lcacheStatus);
it.remove();
}
}
}
// do the deletion asynchronously, after releasing the global lock
Thread cacheFileCleaner =
new Thread(new CacheFileCleanTask(asyncDiskService, FileSystem.getLocal(conf), deleteSet));
cacheFileCleaner.start();
}
/**
* list subfiles
*
* @param hdfsPath !null path probably exists
* @return !null list of enclused file names - emptu if file of not exists
*/
@Override
public String[] ls(final String hdfsPath) {
if (isRunningAsUser()) {
return super.ls(hdfsPath);
}
final FileSystem fs = getDFS();
try {
FileStatus[] statuses = fs.listStatus(new Path(hdfsPath));
if (statuses == null) return new String[0];
List<String> holder = new ArrayList<String>();
for (int i = 0; i < statuses.length; i++) {
FileStatus statuse = statuses[i];
String s = statuse.getPath().getName();
holder.add(s);
}
String[] ret = new String[holder.size()];
holder.toArray(ret);
return ret;
} catch (IOException e) {
throw new RuntimeException(e);
}
}
protected void setup(Context context) throws IOException, InterruptedException {
FileSystem fs = FileSystem.get(context.getConfiguration());
Path cFile = new Path(context.getConfiguration().get("CFILE"));
DataInputStream d = new DataInputStream(fs.open(cFile));
BufferedReader reader = new BufferedReader(new InputStreamReader(d));
String line;
while ((line = reader.readLine()) != null) {
StringTokenizer tokenizer = new StringTokenizer(line.toString());
if (tokenizer.hasMoreTokens()) {
List<Double> centroid = new ArrayList<Double>(58);
while (tokenizer.hasMoreTokens()) {
centroid.add(Double.parseDouble(tokenizer.nextToken()));
}
centroids.add(centroid);
}
}
k = centroids.size();
}
private String runResolveCommand(List<String> args) {
int loopCount = 0;
if (args.size() == 0) {
return null;
}
StringBuffer allOutput = new StringBuffer();
int numProcessed = 0;
if (maxArgs < MIN_ALLOWABLE_ARGS) {
LOG.warn(
"Invalid value "
+ Integer.toString(maxArgs)
+ " for "
+ SCRIPT_ARG_COUNT_KEY
+ "; must be >= "
+ Integer.toString(MIN_ALLOWABLE_ARGS));
return null;
}
while (numProcessed != args.size()) {
int start = maxArgs * loopCount;
List<String> cmdList = new ArrayList<String>();
cmdList.add(scriptName);
for (numProcessed = start;
numProcessed < (start + maxArgs) && numProcessed < args.size();
numProcessed++) {
cmdList.add(args.get(numProcessed));
}
File dir = null;
String userDir;
if ((userDir = System.getProperty("user.dir")) != null) {
dir = new File(userDir);
}
ShellCommandExecutor s = new ShellCommandExecutor(cmdList.toArray(new String[0]), dir);
try {
s.execute();
allOutput.append(s.getOutput() + " ");
} catch (Exception e) {
LOG.warn(StringUtils.stringifyException(e));
return null;
}
loopCount++;
}
return allOutput.toString();
}
public void reduce(IntWritable key, Iterable<Text> values, Context context)
throws IOException, InterruptedException {
List<Double> newCentroid = null;
int numPoints = 0;
for (Text value : values) {
++numPoints;
List<Double> p = GetPoint(value.toString());
if (newCentroid == null) {
// initialize the new centroid to the first element
newCentroid = new ArrayList<Double>(p);
} else {
for (int i = 0; i < newCentroid.size(); i++) {
newCentroid.set(i, newCentroid.get(i) + p.get(i));
}
}
}
// now the newCentroid contains the sum of all the points
// so to get the average of all the points, we need to
// divide each entry by the total number of points
for (int i = 0; i < newCentroid.size(); i++) {
newCentroid.set(i, newCentroid.get(i) / (double) numPoints);
}
// now create a string containing all the new centroid's coordinates
String s = null;
for (Double d : newCentroid) {
if (s == null) {
s = d.toString();
} else {
s += " " + d.toString();
}
}
newCentroids.add(s);
if (newCentroids.size() == 10) {
WriteNewCentroids(context);
}
// output the centroid ID and the centroid data
context.write(key, new Text(s));
}
/**
* Search for pages matching a query, eliminating excessive hits with matching values for a named
* field. Hits after the first <code>maxHitsPerDup</code> are removed from results. The remaining
* hits have {@link Hit#moreFromDupExcluded()} set.
*
* <p>If maxHitsPerDup is zero then all hits are returned.
*
* @param query query
* @param numHits number of requested hits
* @param maxHitsPerDup the maximum hits returned with matching values, or zero
* @param dedupField field name to check for duplicates
* @param sortField Field to sort on (or null if no sorting).
* @param reverse True if we are to reverse sort by <code>sortField</code>.
* @return Hits the matching hits
* @throws IOException
*/
public Hits search(
Query query,
int numHits,
int maxHitsPerDup,
String dedupField,
String sortField,
boolean reverse)
throws IOException {
if (maxHitsPerDup <= 0) // disable dup checking
return search(query, numHits, dedupField, sortField, reverse);
final float rawHitsFactor = this.conf.getFloat("searcher.hostgrouping.rawhits.factor", 2.0f);
int numHitsRaw = (int) (numHits * rawHitsFactor);
if (LOG.isInfoEnabled()) {
LOG.info("searching for " + numHitsRaw + " raw hits");
}
Hits hits = searchBean.search(query, numHitsRaw, dedupField, sortField, reverse);
final long total = hits.getTotal();
final Map<String, DupHits> dupToHits = new HashMap<String, DupHits>();
final List<Hit> resultList = new ArrayList<Hit>();
final Set<Hit> seen = new HashSet<Hit>();
final List<String> excludedValues = new ArrayList<String>();
boolean totalIsExact = true;
for (int rawHitNum = 0; rawHitNum < hits.getTotal(); rawHitNum++) {
// get the next raw hit
if (rawHitNum >= hits.getLength()) {
// optimize query by prohibiting more matches on some excluded values
final Query optQuery = (Query) query.clone();
for (int i = 0; i < excludedValues.size(); i++) {
if (i == MAX_PROHIBITED_TERMS) break;
optQuery.addProhibitedTerm(excludedValues.get(i), dedupField);
}
numHitsRaw = (int) (numHitsRaw * rawHitsFactor);
if (LOG.isInfoEnabled()) {
LOG.info("re-searching for " + numHitsRaw + " raw hits, query: " + optQuery);
}
hits = searchBean.search(optQuery, numHitsRaw, dedupField, sortField, reverse);
if (LOG.isInfoEnabled()) {
LOG.info("found " + hits.getTotal() + " raw hits");
}
rawHitNum = -1;
continue;
}
final Hit hit = hits.getHit(rawHitNum);
if (seen.contains(hit)) continue;
seen.add(hit);
// get dup hits for its value
final String value = hit.getDedupValue();
DupHits dupHits = dupToHits.get(value);
if (dupHits == null) dupToHits.put(value, dupHits = new DupHits());
// does this hit exceed maxHitsPerDup?
if (dupHits.size() == maxHitsPerDup) { // yes -- ignore the hit
if (!dupHits.maxSizeExceeded) {
// mark prior hits with moreFromDupExcluded
for (int i = 0; i < dupHits.size(); i++) {
dupHits.get(i).setMoreFromDupExcluded(true);
}
dupHits.maxSizeExceeded = true;
excludedValues.add(value); // exclude dup
}
totalIsExact = false;
} else { // no -- collect the hit
resultList.add(hit);
dupHits.add(hit);
// are we done?
// we need to find one more than asked for, so that we can tell if
// there are more hits to be shown
if (resultList.size() > numHits) break;
}
}
final Hits results = new Hits(total, resultList.toArray(new Hit[resultList.size()]));
results.setTotalIsExact(totalIsExact);
return results;
}
public void generateFileDetails(JspWriter out, HttpServletRequest req, Configuration conf)
throws IOException, InterruptedException {
int chunkSizeToView = 0;
long startOffset = 0;
int datanodePort;
String blockIdStr = null;
long currBlockId = 0;
blockIdStr = req.getParameter("blockId");
if (blockIdStr == null) {
out.print("Invalid input (blockId absent)");
return;
}
currBlockId = Long.parseLong(blockIdStr);
String datanodePortStr = req.getParameter("datanodePort");
if (datanodePortStr == null) {
out.print("Invalid input (datanodePort absent)");
return;
}
datanodePort = Integer.parseInt(datanodePortStr);
String namenodeInfoPortStr = req.getParameter("namenodeInfoPort");
int namenodeInfoPort = -1;
if (namenodeInfoPortStr != null) namenodeInfoPort = Integer.parseInt(namenodeInfoPortStr);
String chunkSizeToViewStr = req.getParameter("chunkSizeToView");
if (chunkSizeToViewStr != null && Integer.parseInt(chunkSizeToViewStr) > 0) {
chunkSizeToView = Integer.parseInt(chunkSizeToViewStr);
} else {
chunkSizeToView = JspHelper.getDefaultChunkSize(conf);
}
String startOffsetStr = req.getParameter("startOffset");
if (startOffsetStr == null || Long.parseLong(startOffsetStr) < 0) startOffset = 0;
else startOffset = Long.parseLong(startOffsetStr);
String filename = HtmlQuoting.unquoteHtmlChars(req.getParameter("filename"));
if (filename == null || filename.length() == 0) {
out.print("Invalid input");
return;
}
String blockSizeStr = req.getParameter("blockSize");
long blockSize = 0;
if (blockSizeStr == null || blockSizeStr.length() == 0) {
out.print("Invalid input");
return;
}
blockSize = Long.parseLong(blockSizeStr);
String tokenString = req.getParameter(JspHelper.DELEGATION_PARAMETER_NAME);
UserGroupInformation ugi = JspHelper.getUGI(req, conf);
DFSClient dfs = JspHelper.getDFSClient(ugi, jspHelper.nameNodeAddr, conf);
List<LocatedBlock> blocks =
dfs.namenode.getBlockLocations(filename, 0, Long.MAX_VALUE).getLocatedBlocks();
// Add the various links for looking at the file contents
// URL for downloading the full file
String downloadUrl =
"http://"
+ req.getServerName()
+ ":"
+ +req.getServerPort()
+ "/streamFile"
+ URLEncoder.encode(filename, "UTF-8")
+ "?"
+ JspHelper.DELEGATION_PARAMETER_NAME
+ "="
+ tokenString;
out.print("<a name=\"viewOptions\"></a>");
out.print("<a href=\"" + downloadUrl + "\">Download this file</a><br>");
DatanodeInfo chosenNode;
// URL for TAIL
LocatedBlock lastBlk = blocks.get(blocks.size() - 1);
long blockId = lastBlk.getBlock().getBlockId();
try {
chosenNode = jspHelper.bestNode(lastBlk);
} catch (IOException e) {
out.print(e.toString());
dfs.close();
return;
}
String fqdn = InetAddress.getByName(chosenNode.getHost()).getCanonicalHostName();
String tailUrl =
"http://"
+ fqdn
+ ":"
+ chosenNode.getInfoPort()
+ "/tail.jsp?filename="
+ URLEncoder.encode(filename, "UTF-8")
+ "&namenodeInfoPort="
+ namenodeInfoPort
+ "&chunkSizeToView="
+ chunkSizeToView
+ "&referrer="
+ URLEncoder.encode(req.getRequestURL() + "?" + req.getQueryString(), "UTF-8")
+ JspHelper.getDelegationTokenUrlParam(tokenString);
out.print("<a href=\"" + tailUrl + "\">Tail this file</a><br>");
out.print("<form action=\"/browseBlock.jsp\" method=GET>");
out.print("<b>Chunk size to view (in bytes, up to file's DFS block size): </b>");
out.print("<input type=\"hidden\" name=\"blockId\" value=\"" + currBlockId + "\">");
out.print("<input type=\"hidden\" name=\"blockSize\" value=\"" + blockSize + "\">");
out.print("<input type=\"hidden\" name=\"startOffset\" value=\"" + startOffset + "\">");
out.print("<input type=\"hidden\" name=\"filename\" value=\"" + filename + "\">");
out.print("<input type=\"hidden\" name=\"datanodePort\" value=\"" + datanodePort + "\">");
out.print(
"<input type=\"hidden\" name=\"namenodeInfoPort\" value=\"" + namenodeInfoPort + "\">");
out.print(
"<input type=\"text\" name=\"chunkSizeToView\" value="
+ chunkSizeToView
+ " size=10 maxlength=10>");
out.print(" <input type=\"submit\" name=\"submit\" value=\"Refresh\">");
out.print("</form>");
out.print("<hr>");
out.print("<a name=\"blockDetails\"></a>");
out.print("<B>Total number of blocks: " + blocks.size() + "</B><br>");
// generate a table and dump the info
out.println("\n<table>");
for (LocatedBlock cur : blocks) {
out.print("<tr>");
blockId = cur.getBlock().getBlockId();
blockSize = cur.getBlock().getNumBytes();
String blk = "blk_" + Long.toString(blockId);
out.print("<td>" + Long.toString(blockId) + ":</td>");
DatanodeInfo[] locs = cur.getLocations();
for (int j = 0; j < locs.length; j++) {
String datanodeAddr = locs[j].getName();
datanodePort =
Integer.parseInt(
datanodeAddr.substring(datanodeAddr.indexOf(':') + 1, datanodeAddr.length()));
fqdn = InetAddress.getByName(locs[j].getHost()).getCanonicalHostName();
String blockUrl =
"http://"
+ fqdn
+ ":"
+ locs[j].getInfoPort()
+ "/browseBlock.jsp?blockId="
+ Long.toString(blockId)
+ "&blockSize="
+ blockSize
+ "&filename="
+ URLEncoder.encode(filename, "UTF-8")
+ "&datanodePort="
+ datanodePort
+ "&genstamp="
+ cur.getBlock().getGenerationStamp()
+ "&namenodeInfoPort="
+ namenodeInfoPort
+ "&chunkSizeToView="
+ chunkSizeToView;
out.print(
"<td> </td>" + "<td><a href=\"" + blockUrl + "\">" + datanodeAddr + "</a></td>");
}
out.println("</tr>");
}
out.println("</table>");
out.print("<hr>");
String namenodeHost = jspHelper.nameNodeAddr.getHostName();
out.print(
"<br><a href=\"http://"
+ InetAddress.getByName(namenodeHost).getCanonicalHostName()
+ ":"
+ namenodeInfoPort
+ "/dfshealth.jsp\">Go back to DFS home</a>");
dfs.close();
}
public void generateFileChunks(JspWriter out, HttpServletRequest req, Configuration conf)
throws IOException, InterruptedException {
long startOffset = 0;
int datanodePort = 0;
int chunkSizeToView = 0;
String namenodeInfoPortStr = req.getParameter("namenodeInfoPort");
int namenodeInfoPort = -1;
if (namenodeInfoPortStr != null) namenodeInfoPort = Integer.parseInt(namenodeInfoPortStr);
String filename = HtmlQuoting.unquoteHtmlChars(req.getParameter("filename"));
if (filename == null) {
out.print("Invalid input (filename absent)");
return;
}
String blockIdStr = null;
long blockId = 0;
blockIdStr = req.getParameter("blockId");
if (blockIdStr == null) {
out.print("Invalid input (blockId absent)");
return;
}
blockId = Long.parseLong(blockIdStr);
String tokenString = req.getParameter(JspHelper.DELEGATION_PARAMETER_NAME);
UserGroupInformation ugi = JspHelper.getUGI(req, conf);
final DFSClient dfs = JspHelper.getDFSClient(ugi, jspHelper.nameNodeAddr, conf);
Token<BlockTokenIdentifier> accessToken = BlockTokenSecretManager.DUMMY_TOKEN;
if (conf.getBoolean(DFSConfigKeys.DFS_BLOCK_ACCESS_TOKEN_ENABLE_KEY, false)) {
List<LocatedBlock> blks =
dfs.namenode.getBlockLocations(filename, 0, Long.MAX_VALUE).getLocatedBlocks();
if (blks == null || blks.size() == 0) {
out.print("Can't locate file blocks");
dfs.close();
return;
}
for (int i = 0; i < blks.size(); i++) {
if (blks.get(i).getBlock().getBlockId() == blockId) {
accessToken = blks.get(i).getBlockToken();
break;
}
}
}
String blockGenStamp = null;
long genStamp = 0;
blockGenStamp = req.getParameter("genstamp");
if (blockGenStamp == null) {
out.print("Invalid input (genstamp absent)");
return;
}
genStamp = Long.parseLong(blockGenStamp);
String blockSizeStr;
long blockSize = 0;
blockSizeStr = req.getParameter("blockSize");
if (blockSizeStr == null) {
out.print("Invalid input (blockSize absent)");
return;
}
blockSize = Long.parseLong(blockSizeStr);
String chunkSizeToViewStr = req.getParameter("chunkSizeToView");
if (chunkSizeToViewStr != null && Integer.parseInt(chunkSizeToViewStr) > 0)
chunkSizeToView = Integer.parseInt(chunkSizeToViewStr);
else chunkSizeToView = JspHelper.getDefaultChunkSize(conf);
String startOffsetStr = req.getParameter("startOffset");
if (startOffsetStr == null || Long.parseLong(startOffsetStr) < 0) startOffset = 0;
else startOffset = Long.parseLong(startOffsetStr);
String datanodePortStr = req.getParameter("datanodePort");
if (datanodePortStr == null) {
out.print("Invalid input (datanodePort absent)");
return;
}
datanodePort = Integer.parseInt(datanodePortStr);
out.print("<h3>File: ");
JspHelper.printPathWithLinks(
HtmlQuoting.quoteHtmlChars(filename), out, namenodeInfoPort, tokenString);
out.print("</h3><hr>");
String parent = new File(filename).getParent();
JspHelper.printGotoForm(out, namenodeInfoPort, tokenString, HtmlQuoting.quoteHtmlChars(parent));
out.print("<hr>");
out.print(
"<a href=\"http://"
+ req.getServerName()
+ ":"
+ req.getServerPort()
+ "/browseDirectory.jsp?dir="
+ URLEncoder.encode(parent, "UTF-8")
+ "&namenodeInfoPort="
+ namenodeInfoPort
+ "\"><i>Go back to dir listing</i></a><br>");
out.print("<a href=\"#viewOptions\">Advanced view/download options</a><br>");
out.print("<hr>");
// Determine the prev & next blocks
long nextStartOffset = 0;
long nextBlockSize = 0;
String nextBlockIdStr = null;
String nextGenStamp = null;
String nextHost = req.getServerName();
int nextPort = req.getServerPort();
int nextDatanodePort = datanodePort;
// determine data for the next link
if (startOffset + chunkSizeToView >= blockSize) {
// we have to go to the next block from this point onwards
List<LocatedBlock> blocks =
dfs.namenode.getBlockLocations(filename, 0, Long.MAX_VALUE).getLocatedBlocks();
for (int i = 0; i < blocks.size(); i++) {
if (blocks.get(i).getBlock().getBlockId() == blockId) {
if (i != blocks.size() - 1) {
LocatedBlock nextBlock = blocks.get(i + 1);
nextBlockIdStr = Long.toString(nextBlock.getBlock().getBlockId());
nextGenStamp = Long.toString(nextBlock.getBlock().getGenerationStamp());
nextStartOffset = 0;
nextBlockSize = nextBlock.getBlock().getNumBytes();
DatanodeInfo d = jspHelper.bestNode(nextBlock);
String datanodeAddr = d.getName();
nextDatanodePort =
Integer.parseInt(
datanodeAddr.substring(datanodeAddr.indexOf(':') + 1, datanodeAddr.length()));
nextHost = InetAddress.getByName(d.getHost()).getCanonicalHostName();
nextPort = d.getInfoPort();
}
}
}
} else {
// we are in the same block
nextBlockIdStr = blockIdStr;
nextStartOffset = startOffset + chunkSizeToView;
nextBlockSize = blockSize;
nextGenStamp = blockGenStamp;
}
String nextUrl = null;
if (nextBlockIdStr != null) {
nextUrl =
"http://"
+ nextHost
+ ":"
+ nextPort
+ "/browseBlock.jsp?blockId="
+ nextBlockIdStr
+ "&blockSize="
+ nextBlockSize
+ "&startOffset="
+ nextStartOffset
+ "&genstamp="
+ nextGenStamp
+ "&filename="
+ URLEncoder.encode(filename, "UTF-8")
+ "&chunkSizeToView="
+ chunkSizeToView
+ "&datanodePort="
+ nextDatanodePort
+ "&namenodeInfoPort="
+ namenodeInfoPort
+ JspHelper.getDelegationTokenUrlParam(tokenString);
out.print("<a href=\"" + nextUrl + "\">View Next chunk</a> ");
}
// determine data for the prev link
String prevBlockIdStr = null;
String prevGenStamp = null;
long prevStartOffset = 0;
long prevBlockSize = 0;
String prevHost = req.getServerName();
int prevPort = req.getServerPort();
int prevDatanodePort = datanodePort;
if (startOffset == 0) {
List<LocatedBlock> blocks =
dfs.namenode.getBlockLocations(filename, 0, Long.MAX_VALUE).getLocatedBlocks();
for (int i = 0; i < blocks.size(); i++) {
if (blocks.get(i).getBlock().getBlockId() == blockId) {
if (i != 0) {
LocatedBlock prevBlock = blocks.get(i - 1);
prevBlockIdStr = Long.toString(prevBlock.getBlock().getBlockId());
prevGenStamp = Long.toString(prevBlock.getBlock().getGenerationStamp());
prevStartOffset = prevBlock.getBlock().getNumBytes() - chunkSizeToView;
if (prevStartOffset < 0) prevStartOffset = 0;
prevBlockSize = prevBlock.getBlock().getNumBytes();
DatanodeInfo d = jspHelper.bestNode(prevBlock);
String datanodeAddr = d.getName();
prevDatanodePort =
Integer.parseInt(
datanodeAddr.substring(datanodeAddr.indexOf(':') + 1, datanodeAddr.length()));
prevHost = InetAddress.getByName(d.getHost()).getCanonicalHostName();
prevPort = d.getInfoPort();
}
}
}
} else {
// we are in the same block
prevBlockIdStr = blockIdStr;
prevStartOffset = startOffset - chunkSizeToView;
if (prevStartOffset < 0) prevStartOffset = 0;
prevBlockSize = blockSize;
prevGenStamp = blockGenStamp;
}
String prevUrl = null;
if (prevBlockIdStr != null) {
prevUrl =
"http://"
+ prevHost
+ ":"
+ prevPort
+ "/browseBlock.jsp?blockId="
+ prevBlockIdStr
+ "&blockSize="
+ prevBlockSize
+ "&startOffset="
+ prevStartOffset
+ "&filename="
+ URLEncoder.encode(filename, "UTF-8")
+ "&chunkSizeToView="
+ chunkSizeToView
+ "&genstamp="
+ prevGenStamp
+ "&datanodePort="
+ prevDatanodePort
+ "&namenodeInfoPort="
+ namenodeInfoPort
+ JspHelper.getDelegationTokenUrlParam(tokenString);
out.print("<a href=\"" + prevUrl + "\">View Prev chunk</a> ");
}
out.print("<hr>");
out.print("<textarea cols=\"100\" rows=\"25\" wrap=\"virtual\" style=\"width:100%\" READONLY>");
try {
jspHelper.streamBlockInAscii(
new InetSocketAddress(req.getServerName(), datanodePort),
blockId,
accessToken,
genStamp,
blockSize,
startOffset,
chunkSizeToView,
out,
conf);
} catch (Exception e) {
out.print(e);
}
out.print("</textarea>");
dfs.close();
}
public static List<String> getText(BytesWritable value, Boolean tokenizep)
throws InterruptedException {
Session s = Session.getDefaultInstance(new Properties());
InputStream is = new ByteArrayInputStream(value.getBytes());
List<String> out = new ArrayList<String>();
try {
MimeMessage message = new MimeMessage(s, is);
message.getAllHeaderLines();
Analyzer standard_analyzer = new StandardAnalyzer(Version.LUCENE_43);
Analyzer email_analyzer = new UAX29URLEmailAnalyzer(Version.LUCENE_43);
Address[] fromAddrs = message.getFrom();
String fromAddrstr = "";
if (fromAddrs != null) {
for (Address addr : fromAddrs) {
fromAddrstr += (addr.toString() + " ");
}
}
Address[] toAddrs = message.getAllRecipients();
String toAddrstr = "";
if (toAddrs != null) {
for (Address addr : toAddrs) {
toAddrstr += (addr.toString() + " ");
}
}
String subject = message.getSubject();
String body = "";
try {
Object content = message.getContent();
// System.err.println(content.getContentType());
if (content instanceof String) {
body = (String) content;
} else if (content instanceof Multipart) {
Multipart mp = (Multipart) content;
for (int i = 0; i < mp.getCount(); i++) {
BodyPart bp = mp.getBodyPart(i);
// System.err.println(bp.getContentType());
Object c = bp.getContent();
if (c instanceof String) {
body = (String) c;
}
}
}
// people do really evil things with email, we're not sorting through it all now
} catch (DecodingException e) {
System.err.println("DecodingException");
} catch (UnsupportedEncodingException e) {
System.err.println("UnsuportedEncodingException");
} catch (IOException e) {
System.err.println("IOException");
}
if (tokenizep) {
List<String> fromData = new ArrayList<String>();
List<String> toData = new ArrayList<String>();
List<String> subjectData = new ArrayList<String>();
List<String> bodyData = new ArrayList<String>();
if (fromAddrstr != null) {
fromData = tokenizeString(email_analyzer, fromAddrstr);
}
if (toAddrstr != null) {
toData = tokenizeString(email_analyzer, toAddrstr);
}
if (subject != null) {
subjectData = tokenizeString(standard_analyzer, subject);
}
if (body != null) {
bodyData = tokenizeString(standard_analyzer, body);
}
out.add("FROM ");
out.addAll(fromData);
out.add("TO ");
out.addAll(toData);
out.add("SUBJECT ");
out.addAll(subjectData);
out.add("BODY ");
out.addAll(bodyData);
} else {
// if not tokenizep, return list with from and subject fields only
out.add(fromAddrstr);
out.add(subject);
}
} catch (MessagingException e) {
System.err.println("MessagineException");
}
return out;
}
public void generateFileChunks(JspWriter out, HttpServletRequest req) throws IOException {
long startOffset = 0;
int chunkSizeToView = 0;
String referrer = req.getParameter("referrer");
boolean noLink = false;
if (referrer == null) {
noLink = true;
}
String filename = req.getParameter("filename");
if (filename == null) {
out.print("Invalid input (file name absent)");
return;
}
String namenodeInfoPortStr = req.getParameter("namenodeInfoPort");
int namenodeInfoPort = -1;
if (namenodeInfoPortStr != null) namenodeInfoPort = Integer.parseInt(namenodeInfoPortStr);
String chunkSizeToViewStr = req.getParameter("chunkSizeToView");
if (chunkSizeToViewStr != null && Integer.parseInt(chunkSizeToViewStr) > 0)
chunkSizeToView = Integer.parseInt(chunkSizeToViewStr);
else chunkSizeToView = jspHelper.defaultChunkSizeToView;
if (!noLink) {
out.print("<h3>Tail of File: ");
JspHelper.printPathWithLinks(filename, out, namenodeInfoPort);
out.print("</h3><hr>");
out.print("<a href=\"" + referrer + "\">Go Back to File View</a><hr>");
} else {
out.print("<h3>" + filename + "</h3>");
}
out.print("<b>Chunk size to view (in bytes, up to file's DFS block size): </b>");
out.print(
"<input type=\"text\" name=\"chunkSizeToView\" value="
+ chunkSizeToView
+ " size=10 maxlength=10>");
out.print(" <input type=\"submit\" name=\"submit\" value=\"Refresh\"><hr>");
out.print("<input type=\"hidden\" name=\"filename\" value=\"" + filename + "\">");
out.print(
"<input type=\"hidden\" name=\"namenodeInfoPort\" value=\"" + namenodeInfoPort + "\">");
if (!noLink) out.print("<input type=\"hidden\" name=\"referrer\" value=\"" + referrer + "\">");
// fetch the block from the datanode that has the last block for this file
DFSClient dfs = new DFSClient(jspHelper.nameNodeAddr, jspHelper.conf);
List<LocatedBlock> blocks =
dfs.namenode.getBlockLocations(filename, 0, Long.MAX_VALUE).getLocatedBlocks();
if (blocks == null || blocks.size() == 0) {
out.print("No datanodes contain blocks of file " + filename);
dfs.close();
return;
}
LocatedBlock lastBlk = blocks.get(blocks.size() - 1);
long blockSize = lastBlk.getBlock().getNumBytes();
long blockId = lastBlk.getBlock().getBlockId();
long genStamp = lastBlk.getBlock().getGenerationStamp();
DatanodeInfo chosenNode;
try {
chosenNode = jspHelper.bestNode(lastBlk);
} catch (IOException e) {
out.print(e.toString());
dfs.close();
return;
}
InetSocketAddress addr = NetUtils.createSocketAddr(chosenNode.getName());
// view the last chunkSizeToView bytes while Tailing
if (blockSize >= chunkSizeToView) startOffset = blockSize - chunkSizeToView;
else startOffset = 0;
out.print("<textarea cols=\"100\" rows=\"25\" wrap=\"virtual\" style=\"width:100%\" READONLY>");
jspHelper.streamBlockInAscii(
addr, blockId, genStamp, blockSize, startOffset, chunkSizeToView, out);
out.print("</textarea>");
dfs.close();
}
public int run(String[] args) throws Exception {
// printUsage();
/*
* SETUP
*/
Configuration argConf = getConf();
Hashtable<String, String> confArg = new Hashtable<String, String>();
setup(confArg, argConf);
Date currentTime = new Date();
Date endDate = new Date(new Long(confArg.get("timestamp_stop")));
Boolean full_run = confArg.get("intermediate").matches("(?i).*true.*");
Boolean quick_add = confArg.get("quick_add").matches("(?i).*true.*");
logger.info("Running GeStore");
// ZooKeeper setup
Configuration config = HBaseConfiguration.create();
zkWatcher = new ZooKeeperWatcher(config, "Testing", new HBaseAdmin(config));
zkInstance =
new ZooKeeper(
ZKConfig.getZKQuorumServersString(config),
config.getInt("zookeeper.session.timeout", -1),
zkWatcher);
if (!confArg.get("task_id").isEmpty()) {
confArg.put("temp_path", confArg.get("temp_path") + confArg.get("task_id"));
}
String lockRequest = confArg.get("file_id");
if (!confArg.get("run_id").isEmpty())
lockRequest = lockRequest + "_" + confArg.get("run_id") + "_";
if (!confArg.get("task_id").isEmpty())
lockRequest = lockRequest + "_" + confArg.get("task_id") + "_";
// Get type of movement
toFrom type_move = checkArgs(confArg);
if (type_move == toFrom.LOCAL2REMOTE && !confArg.get("format").equals("unknown")) {
List<String> arguments = new ArrayList<String>();
arguments.add("-Dinput=" + confArg.get("local_path"));
arguments.add("-Dtable=" + confArg.get("file_id"));
arguments.add("-Dtimestamp=" + confArg.get("timestamp_stop"));
arguments.add("-Dtype=" + confArg.get("format"));
arguments.add("-Dtarget_dir=" + confArg.get("base_path") + "_" + confArg.get("file_id"));
arguments.add("-Dtemp_hdfs_path=" + confArg.get("temp_path"));
arguments.add("-Drun_id=" + confArg.get("run_id"));
if (!confArg.get("run_id").isEmpty()) arguments.add("-Drun_id=" + confArg.get("run_id"));
if (!confArg.get("task_id").isEmpty()) arguments.add("-Dtask_id=" + confArg.get("task_id"));
if (quick_add) arguments.add("-Dquick_add=" + confArg.get("quick_add"));
String lockName = lock(lockRequest);
String[] argumentString = arguments.toArray(new String[arguments.size()]);
adddb.main(argumentString);
unlock(lockName);
System.exit(0);
}
// Database registration
dbutil db_util = new dbutil(config);
db_util.register_database(confArg.get("db_name_files"), true);
db_util.register_database(confArg.get("db_name_runs"), true);
db_util.register_database(confArg.get("db_name_updates"), true);
FileSystem hdfs = FileSystem.get(config);
FileSystem localFS = FileSystem.getLocal(config);
// Get source type
confArg.put("source", getSource(db_util, confArg.get("db_name_files"), confArg.get("file_id")));
confArg.put(
"database", isDatabase(db_util, confArg.get("db_name_files"), confArg.get("file_id")));
if (!confArg.get("source").equals("local")
&& type_move == toFrom.REMOTE2LOCAL
&& !confArg.get("timestamp_stop").equals(Integer.toString(Integer.MAX_VALUE))) {
confArg.put("timestamp_stop", Long.toString(latestVersion(confArg, db_util)));
}
/*
* Get previous timestamp
*/
Get run_id_get = new Get(confArg.get("run_id").getBytes());
Result run_get = db_util.doGet(confArg.get("db_name_runs"), run_id_get);
KeyValue run_file_prev =
run_get.getColumnLatest(
"d".getBytes(), (confArg.get("file_id") + "_db_timestamp").getBytes());
String last_timestamp = new String("0");
if (null != run_file_prev && !confArg.get("source").equals("local")) {
long last_timestamp_real = run_file_prev.getTimestamp();
Long current_timestamp = new Long(confArg.get("timestamp_real"));
if ((current_timestamp - last_timestamp_real) > 36000) {
last_timestamp = new String(run_file_prev.getValue());
Integer lastTimestamp = new Integer(last_timestamp);
lastTimestamp += 1;
last_timestamp = lastTimestamp.toString();
logger.info("Last timestamp: " + last_timestamp + " End data: " + endDate);
Date last_run = new Date(run_file_prev.getTimestamp());
if (last_run.before(endDate) && !full_run) {
confArg.put("timestamp_start", last_timestamp);
}
}
}
Integer tse = new Integer(confArg.get("timestamp_stop"));
Integer tss = new Integer(confArg.get("timestamp_start"));
if (tss > tse) {
logger.info("No new version of requested file.");
return 0;
}
/*
* Generate file
*/
String lockName = lock(lockRequest);
Get file_id_get = new Get(confArg.get("file_id").getBytes());
Result file_get = db_util.doGet(confArg.get("db_name_files"), file_id_get);
if (!file_get.isEmpty()) {
boolean found =
hasFile(
db_util,
hdfs,
confArg.get("db_name_files"),
confArg.get("file_id"),
getFullPath(confArg));
if (confArg.get("source").equals("fullfile")) {
found = false;
}
String filenames_put =
getFileNames(
db_util, confArg.get("db_name_files"), confArg.get("file_id"), getFullPath(confArg));
// Filename not found in file database
if (!found && type_move == toFrom.REMOTE2LOCAL) {
if (!confArg.get("source").equals("local")) {
// Generate intermediate file
if (getFile(hdfs, confArg, db_util) == null) {
unlock(lockName);
return 1;
}
// Put generated file into file database
if (!confArg.get("format").equals("fullfile")) {
putFileEntry(
db_util,
hdfs,
confArg.get("db_name_files"),
confArg.get("file_id"),
confArg.get("full_file_name"),
confArg.get("source"));
}
} else {
logger.warn("Remote file not found, and cannot be generated! File: " + confArg);
unlock(lockName);
return 1;
}
}
} else {
if (type_move == toFrom.REMOTE2LOCAL) {
logger.warn("Remote file not found, and cannot be generated.");
unlock(lockName);
return 1;
}
}
/*
* Copy file
* Update tables
*/
if (type_move == toFrom.LOCAL2REMOTE) {
if (!confArg.get("format").equals("fullfile")) {
putFileEntry(
db_util,
hdfs,
confArg.get("db_name_files"),
confArg.get("file_id"),
getFullPath(confArg),
confArg.get("source"));
}
putRunEntry(
db_util,
confArg.get("db_name_runs"),
confArg.get("run_id"),
confArg.get("file_id"),
confArg.get("type"),
confArg.get("timestamp_real"),
confArg.get("timestamp_stop"),
getFullPath(confArg),
confArg.get("delimiter"));
hdfs.copyFromLocalFile(new Path(confArg.get("local_path")), new Path(getFullPath(confArg)));
} else if (type_move == toFrom.REMOTE2LOCAL) {
FileStatus[] files = hdfs.globStatus(new Path(getFullPath(confArg) + "*"));
putRunEntry(
db_util,
confArg.get("db_name_runs"),
confArg.get("run_id"),
confArg.get("file_id"),
confArg.get("type"),
confArg.get("timestamp_real"),
confArg.get("timestamp_stop"),
getFullPath(confArg),
confArg.get("delimiter"));
unlock(lockName);
for (FileStatus file : files) {
Path cur_file = file.getPath();
Path cur_local_path =
new Path(new String(confArg.get("local_path") + confArg.get("file_id")));
String suffix = getSuffix(getFileName(confArg), cur_file.getName());
if (suffix.length() > 0) {
cur_local_path = cur_local_path.suffix(new String("." + suffix));
}
if (confArg.get("copy").equals("true")) {
String crc = hdfs.getFileChecksum(cur_file).toString();
if (checksumLocalTest(cur_local_path, crc)) {
continue;
} else {
hdfs.copyToLocalFile(cur_file, cur_local_path);
writeChecksum(cur_local_path, crc);
}
} else {
System.out.println(cur_local_path + "\t" + cur_file);
}
}
}
unlock(lockName);
return 0;
}