yacy_search_server/htroot/yacy/transferRWI.java

220 lines
11 KiB
Java
Raw Normal View History

// transferRWI.java
// -----------------------
// part of the AnomicHTTPD caching proxy
// (C) by Michael Peter Christen; mc@yacy.net
// first published on http://www.anomic.de
// Frankfurt, Germany, 2004, 2005
//
// $LastChangedDate$
// $LastChangedRevision$
// $LastChangedBy$
//
// This program is free software; you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation; either version 2 of the License, or
// (at your option) any later version.
//
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License for more details.
//
// You should have received a copy of the GNU General Public License
// along with this program; if not, write to the Free Software
// Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
// You must compile this file with
// javac -classpath .:../classes transferRWI.java
*) Asynchronous queuing of crawl job URLs (stackCrawl) various checks like the blacklist check or the robots.txt disallow check are now done by a separate thread to unburden the indexer thread(s) TODO: maybe we have to introduce a threadpool here if it turn out that this single thread is a bottleneck because of the time consuming robots.txt downloads *) improved index transfer The index selection and transmission is done in parallel now to improve index transfer performance. TODO: maybe we could speed up performance by unsing multiple transmission threads in parallel instead of only a single one. *) gzip encoded post requests it is now configureable if a gzip encoded post request should be send on intex transfer/distribution *) storage Peer (very experimentell and not optimized yet) Now it's possible to send the result of the yacy indexer thread to a remote peer istead of storing the indexed words locally. This could be done by setting the property "storagePeerHash" in the yacy config file - Please note that if the index transfer fails, the index ist stored locally. - TODO: currently this index transfer is done by the indexer thread. To seedup the indexer a) this transmission should be done in parallel and b) multiple chunks should be bundled and transfered together *) general performance improvements - better memory cleanup after http request processing has finished - replacing some string concatenations with stringBuffers - replacing BufferedInputStreams with serverByteBuffer - replacing vectors with arraylists wherever possible - replacing hashtables with hashmaps wherever possible This was done because function calls to verctor or hashtable functions take 3 time longer than calls to functions of arraylists or hashmaps. TODO: we should take a look on the class serverObject which is inherited from hashmap Do we realy need a synchronization for this class? TODO: replace arraylists with linkedLists if random access to the list elements is not needed *) Robots Parser supports if-modified-since downloads now If the downloaded robots.txt file is older than 7 days the robots parser tries to download the robots.txt with the if-modified-since header to avoid unnecessary downloads if the file was not changed. Additionally the ETag header is used to detect changes. *) Crawler: better handling of unsupported mimeTypes + FileExtension *) Bugfix: plasmaWordIndexEntity was not closed correctly in - query.java - plasmaswitchboard.java *) function minimizeUrlDB added to yacy.java this function tests the current urlHashDB for unused urls ATTENTION: please don't use this function at the moment because it causes the wordIndexDB to flush all words into the word directory! git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@853 6c8d7289-2bf4-0310-a012-ef5d649a1542
2005-10-05 12:45:33 +02:00
import java.io.IOException;
import java.util.HashSet;
import java.util.Iterator;
import java.util.List;
import de.anomic.data.Blacklist;
import de.anomic.http.httpRequestHeader;
import de.anomic.kelondro.text.referencePrototype.WordReferenceRow;
import de.anomic.kelondro.util.FileUtils;
import de.anomic.kelondro.util.Log;
import de.anomic.plasma.plasmaSwitchboard;
import de.anomic.plasma.plasmaSwitchboardConstants;
import de.anomic.server.serverCore;
import de.anomic.server.serverObjects;
import de.anomic.server.serverSwitch;
import de.anomic.xml.RSSFeed;
import de.anomic.xml.RSSMessage;
import de.anomic.yacy.yacyCore;
import de.anomic.yacy.yacyNetwork;
import de.anomic.yacy.yacySeed;
replaced old DHT transmission method with new method. Many things have changed! some of them: - after a index selection is made, the index is splitted into its vertical components - from differrent index selctions the splitted components can be accumulated before they are placed into the transmission queue - each splitted chunk gets its own transmission thread - multiple transmission threads are started concurrently - the process can be monitored with the blocking queue servlet To implement that, a new package de.anomic.yacy.dht was created. Some old files have been removed. The new index distribution model using a vertical DHT was implemented. An abstraction of this model is implemented in the new dht package as interface. The freeworld network has now a configuration of two vertial partitions; sixteen partitions are planned and will be configured if the process is bug-free. This modification has three main targets: - enhance the DHT transmission speed - with a vertical DHT, a search will speed up. With two partitions, two times. With sixteen, sixteen times. - the vertical DHT will apply a semi-dht for URLs, and peers will receive a fraction of the overall URLs they received before. with two partitions, the fractions will be halve. With sixteen partitions, a 1/16 of the previous number of URLs. BE CAREFULL, THIS IS A MAJOR CODE CHANGE, POSSIBLY FULL OF BUGS AND HARMFUL THINGS. git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@5586 6c8d7289-2bf4-0310-a012-ef5d649a1542
2009-02-10 01:06:59 +01:00
import de.anomic.yacy.dht.FlatWordPartitionScheme;
*) Asynchronous queuing of crawl job URLs (stackCrawl) various checks like the blacklist check or the robots.txt disallow check are now done by a separate thread to unburden the indexer thread(s) TODO: maybe we have to introduce a threadpool here if it turn out that this single thread is a bottleneck because of the time consuming robots.txt downloads *) improved index transfer The index selection and transmission is done in parallel now to improve index transfer performance. TODO: maybe we could speed up performance by unsing multiple transmission threads in parallel instead of only a single one. *) gzip encoded post requests it is now configureable if a gzip encoded post request should be send on intex transfer/distribution *) storage Peer (very experimentell and not optimized yet) Now it's possible to send the result of the yacy indexer thread to a remote peer istead of storing the indexed words locally. This could be done by setting the property "storagePeerHash" in the yacy config file - Please note that if the index transfer fails, the index ist stored locally. - TODO: currently this index transfer is done by the indexer thread. To seedup the indexer a) this transmission should be done in parallel and b) multiple chunks should be bundled and transfered together *) general performance improvements - better memory cleanup after http request processing has finished - replacing some string concatenations with stringBuffers - replacing BufferedInputStreams with serverByteBuffer - replacing vectors with arraylists wherever possible - replacing hashtables with hashmaps wherever possible This was done because function calls to verctor or hashtable functions take 3 time longer than calls to functions of arraylists or hashmaps. TODO: we should take a look on the class serverObject which is inherited from hashmap Do we realy need a synchronization for this class? TODO: replace arraylists with linkedLists if random access to the list elements is not needed *) Robots Parser supports if-modified-since downloads now If the downloaded robots.txt file is older than 7 days the robots parser tries to download the robots.txt with the if-modified-since header to avoid unnecessary downloads if the file was not changed. Additionally the ETag header is used to detect changes. *) Crawler: better handling of unsupported mimeTypes + FileExtension *) Bugfix: plasmaWordIndexEntity was not closed correctly in - query.java - plasmaswitchboard.java *) function minimizeUrlDB added to yacy.java this function tests the current urlHashDB for unused urls ATTENTION: please don't use this function at the moment because it causes the wordIndexDB to flush all words into the word directory! git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@853 6c8d7289-2bf4-0310-a012-ef5d649a1542
2005-10-05 12:45:33 +02:00
public final class transferRWI {
public static serverObjects respond(final httpRequestHeader header, final serverObjects post, final serverSwitch<?> env) throws InterruptedException {
// return variable that accumulates replacements
final plasmaSwitchboard sb = (plasmaSwitchboard) env;
final serverObjects prop = new serverObjects();
final String contentType = header.getContentType();
if ((post == null) || (env == null)) {
logWarning(contentType, "post or env is null!");
return prop;
}
if (!yacyNetwork.authentifyRequest(post, env)) {
logWarning(contentType, "not authentified");
return prop;
}
if (!post.containsKey("wordc")) {
logWarning(contentType, "missing wordc");
return prop;
}
if (!post.containsKey("entryc")) {
logWarning(contentType, "missing entryc");
return prop;
}
// request values
final String iam = post.get("iam", ""); // seed hash of requester
final String youare = post.get("youare", ""); // seed hash of the target peer, needed for network stability
// final String key = (String) post.get("key", ""); // transmission key
final int wordc = post.getInt("wordc", 0); // number of different words
final int entryc = post.getInt("entryc", 0); // number of entries in indexes
byte[] indexes = post.get("indexes", "").getBytes(); // the indexes, as list of word entries
boolean granted = sb.getConfig("allowReceiveIndex", "false").equals("true");
final boolean blockBlacklist = sb.getConfig("indexReceiveBlockBlacklist", "false").equals("true");
final long cachelimit = sb.getConfigLong(plasmaSwitchboardConstants.WORDCACHE_MAX_COUNT, 100000);
final yacySeed otherPeer = sb.webIndex.peers().get(iam);
final String otherPeerName = iam + ":" + ((otherPeer == null) ? "NULL" : (otherPeer.getName() + "/" + otherPeer.getVersion()));
// response values
int pause = 0;
String result = "ok";
final StringBuilder unknownURLs = new StringBuilder();
if ((youare == null) || (!youare.equals(sb.webIndex.peers().mySeed().hash))) {
sb.getLog().logInfo("Rejecting RWIs from peer " + otherPeerName + ". Wrong target. Wanted peer=" + youare + ", iam=" + sb.webIndex.peers().mySeed().hash);
result = "wrong_target";
pause = 0;
} else if ((!granted) || (sb.isRobinsonMode())) {
// we dont want to receive indexes
sb.getLog().logInfo("Rejecting RWIs from peer " + otherPeerName + ". Not granted.");
result = "not_granted";
pause = 0;
} else if (sb.webIndex.index().getBufferSize() > cachelimit) {
// we are too busy to receive indexes
sb.getLog().logInfo("Rejecting RWIs from peer " + otherPeerName + ". We are too busy (buffersize=" + sb.webIndex.index().getBufferSize() + ").");
granted = false; // don't accept more words if there are too many words to flush
result = "busy";
pause = 60000;
} else {
// we want and can receive indexes
// log value status (currently added to find outOfMemory error
if (sb.getLog().isFine()) sb.getLog().logFine("Processing " + indexes.length + " bytes / " + wordc + " words / " + entryc + " entries from " + otherPeerName);
final long startProcess = System.currentTimeMillis();
// decode request
final List<String> v = FileUtils.strings(indexes, null);
// free memory
indexes = null;
// the value-vector should now have the same length as entryc
if (v.size() != entryc) sb.getLog().logSevere("ERROR WITH ENTRY COUNTER: v=" + v.size() + ", entryc=" + entryc);
// now parse the Strings in the value-vector and write index entries
String estring;
int p;
String wordHash;
String urlHash;
WordReferenceRow iEntry;
final HashSet<String> unknownURL = new HashSet<String>();
final HashSet<String> knownURL = new HashSet<String>();
final String[] wordhashes = new String[v.size()];
int received = 0;
int blocked = 0;
int receivedURL = 0;
final Iterator<String> i = v.iterator();
while (i.hasNext()) {
serverCore.checkInterruption();
estring = i.next();
// check if RWI entry is well-formed
p = estring.indexOf("{");
if ((p < 0) || (estring.indexOf("x=") < 0)) {
blocked++;
continue;
}
wordHash = estring.substring(0, p);
wordhashes[received] = wordHash;
iEntry = new WordReferenceRow(estring.substring(p));
urlHash = iEntry.metadataHash();
// block blacklisted entries
if ((blockBlacklist) && (plasmaSwitchboard.urlBlacklist.hashInBlacklistedCache(Blacklist.BLACKLIST_DHT, urlHash))) {
if (yacyCore.log.isFine()) yacyCore.log.logFine("transferRWI: blocked blacklisted URLHash '" + urlHash + "' from peer " + otherPeerName);
blocked++;
continue;
}
// learn entry
try {
sb.webIndex.index().add(wordHash, iEntry);
} catch (IOException e) {
e.printStackTrace();
}
serverCore.checkInterruption();
// check if we need to ask for the corresponding URL
if (!(knownURL.contains(urlHash)||unknownURL.contains(urlHash))) try {
if (sb.webIndex.metadata().exists(urlHash)) {
knownURL.add(urlHash);
} else {
unknownURL.add(urlHash);
}
receivedURL++;
} catch (final Exception ex) {
sb.getLog().logWarning(
"transferRWI: DB-Error while trying to determine if URL with hash '" +
urlHash + "' is known.", ex);
}
received++;
}
sb.webIndex.peers().mySeed().incRI(received);
// finally compose the unknownURL hash list
final Iterator<String> it = unknownURL.iterator();
unknownURLs.ensureCapacity(unknownURL.size() * 25);
while (it.hasNext()) {
unknownURLs.append(",").append(it.next());
}
if (unknownURLs.length() > 0) { unknownURLs.delete(0, 1); }
if ((wordhashes.length == 0) || (received == 0)) {
sb.getLog().logInfo("Received 0 RWIs from " + otherPeerName + ", processed in " + (System.currentTimeMillis() - startProcess) + " milliseconds, requesting " + unknownURL.size() + " URLs, blocked " + blocked + " RWIs");
} else {
final long avdist = (FlatWordPartitionScheme.std.dhtDistance(wordhashes[0], null, sb.webIndex.peers().mySeed()) + FlatWordPartitionScheme.std.dhtDistance(wordhashes[received - 1], null, sb.webIndex.peers().mySeed())) / 2;
sb.getLog().logInfo("Received " + received + " Entries " + wordc + " Words [" + wordhashes[0] + " .. " + wordhashes[received - 1] + "]/" + avdist + " from " + otherPeerName + ", processed in " + (System.currentTimeMillis() - startProcess) + " milliseconds, requesting " + unknownURL.size() + "/" + receivedURL + " URLs, blocked " + blocked + " RWIs");
RSSFeed.channels(RSSFeed.INDEXRECEIVE).addMessage(new RSSMessage("Received " + received + " RWIs [" + wordhashes[0] + " .. " + wordhashes[received - 1] + "]/" + avdist + " from " + otherPeerName + ", requesting " + unknownURL.size() + " URLs, blocked " + blocked, "", ""));
}
result = "ok";
pause = (int) (sb.webIndex.index().getBufferSize() * 20000 / sb.getConfigLong(plasmaSwitchboardConstants.WORDCACHE_MAX_COUNT, 100000)); // estimation of necessary pause time
}
* Complete number localization and provide a more reasonable interface to serverObjects: - put(key, value) methods are now used if a value added to the map should be kept as it is. Numbers are transformed (but not formatted) to an equivalent String representation. - putASIS(...) have been removed, now done with simple put(...) (see above). - puNum(...) can be used for number values which should be stored in a formatted way, either depending on the current locale setting for yacy (default) or in a "none" locale (see javadocs and setLocalize()). - putHTML(...) escapes special characters into corresponding HTML enities ('<' => '&lt;') which was done with put(...) before and so was called too often, becauses it is necessary only for very few cases. Additionally there is a "forXML" mode which only replaces < > & ". In short: Use put(...) for almost everything, use putXY(...) if you need some special transformation of the value. A few bugs have been fixed as well, and there should be a small performance improvement for complex pages with a lot of values. * added additional Sum/Avg rows to access tracker pages, see http://forum.yacy-websuche.de/viewtopic.php?f=5&t=456 * removed duplicate code (mostly related to the big changes above). TODO: - make sure, number formats work as expected _everywhere_, report overseen stuff http://forum.yacy-websuche.de/viewtopic.php?f=5&t=437 - probably a good idea to add special putDate() methods as they are used in many pages and create duplicated formatting code + maybe some centralized handling for memory value formatting. - further improve the speed of page creation for the WatchCrawler. git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@4178 6c8d7289-2bf4-0310-a012-ef5d649a1542
2007-10-24 23:38:19 +02:00
prop.put("unknownURL", unknownURLs.toString());
prop.put("result", result);
prop.put("pause", pause);
// return rewrite properties
return prop;
}
/**
* @param requestIdentifier
* @param msg
*/
private static void logWarning(final String requestIdentifier, final String msg) {
Log.logWarning("transferRWI", requestIdentifier +" "+ msg);
}
}