mirror of
https://github.com/yacy/yacy_search_server.git
synced 2024-09-21 00:00:13 +02:00
c3e7efa846
saves memory and speeds up enqueueContainers by limiting the size of transfer.Chunk saves network bandwidth by not transmitting RWIs that would get discarded at the target anyway git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7995 6c8d7289-2bf4-0310-a012-ef5d649a1542
309 lines
13 KiB
Java
309 lines
13 KiB
Java
// Transmission.java
|
|
// (C) 2009 by Michael Peter Christen; mc@yacy.net, Frankfurt a. M., Germany
|
|
// first published 09.02.2009 on http://yacy.net
|
|
//
|
|
// $LastChangedDate$
|
|
// $LastChangedRevision$
|
|
// $LastChangedBy$
|
|
//
|
|
// LICENSE
|
|
//
|
|
// 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
|
|
|
|
package net.yacy.peers.dht;
|
|
|
|
import java.util.ArrayList;
|
|
import java.util.Iterator;
|
|
import java.util.Random;
|
|
import java.util.TreeMap;
|
|
|
|
import net.yacy.cora.document.ASCII;
|
|
import net.yacy.kelondro.data.meta.URIMetadataRow;
|
|
import net.yacy.kelondro.data.word.Word;
|
|
import net.yacy.kelondro.data.word.WordReference;
|
|
import net.yacy.kelondro.data.word.WordReferenceRow;
|
|
import net.yacy.kelondro.index.HandleSet;
|
|
import net.yacy.kelondro.index.RowSpaceExceededException;
|
|
import net.yacy.kelondro.logging.Log;
|
|
import net.yacy.kelondro.order.Base64Order;
|
|
import net.yacy.kelondro.rwi.ReferenceContainer;
|
|
import net.yacy.kelondro.rwi.ReferenceContainerCache;
|
|
import net.yacy.kelondro.workflow.WorkflowJob;
|
|
import net.yacy.peers.Protocol;
|
|
import net.yacy.peers.Seed;
|
|
import net.yacy.peers.SeedDB;
|
|
import net.yacy.search.index.Segment;
|
|
|
|
import java.util.List;
|
|
import java.util.SortedMap;
|
|
|
|
public class Transmission {
|
|
|
|
// The number of RWIs we can be sure a remote peer will accept
|
|
// anything beyond that might get discarded without notice
|
|
public static final int maxRWIsCount = 1000; // since SVN 7993 hardcoded in htroot/yacy/transferRWI.java:161
|
|
|
|
protected Log log;
|
|
protected Segment segment;
|
|
protected SeedDB seeds;
|
|
protected boolean gzipBody4Transfer;
|
|
protected int timeout4Transfer;
|
|
|
|
public Transmission(
|
|
Log log,
|
|
Segment segment,
|
|
SeedDB seeds,
|
|
boolean gzipBody4Transfer,
|
|
int timeout4Transfer) {
|
|
this.log = log;
|
|
this.segment = segment;
|
|
this.seeds = seeds;
|
|
this.gzipBody4Transfer = gzipBody4Transfer;
|
|
this.timeout4Transfer = timeout4Transfer;
|
|
}
|
|
|
|
public Chunk newChunk(byte[] primaryTarget, final List<Seed> targets) {
|
|
return new Chunk(primaryTarget, targets);
|
|
}
|
|
|
|
public class Chunk extends WorkflowJob implements Iterable<ReferenceContainer<WordReference>> {
|
|
/**
|
|
* a dispatcher entry contains
|
|
* - the primary target, which is a word hash, as marker for the entry
|
|
* - a set of indexContainers in a cache,
|
|
* - the associated URLs in a set to have a cache for the transmission
|
|
* to multiple peers and to ensure that all entries in the indexContainers
|
|
* have a reference in the urls
|
|
* - a set of yacy seeds which will shrink as the containers are transmitted to them
|
|
* - a counter that gives the number of sucessful and unsuccessful transmissions so far
|
|
*/
|
|
private final byte[] primaryTarget;
|
|
private final ReferenceContainerCache<WordReference> containers;
|
|
private final SortedMap<byte[], URIMetadataRow> references;
|
|
private final HandleSet badReferences;
|
|
private final List<Seed> targets;
|
|
private int hit, miss;
|
|
|
|
/**
|
|
* generate a new dispatcher target. such a target is defined with a primary target and
|
|
* a set of target peers that shall receive the entries of the containers
|
|
* the payloadrow defines the structure of container entries
|
|
* @param primaryTarget
|
|
* @param targets
|
|
*/
|
|
public Chunk(byte[] primaryTarget, final List<Seed> targets) {
|
|
super();
|
|
this.primaryTarget = primaryTarget;
|
|
this.containers = new ReferenceContainerCache<WordReference>(Segment.wordReferenceFactory, Segment.wordOrder, Word.commonHashLength);
|
|
this.references = new TreeMap<byte[], URIMetadataRow>(Base64Order.enhancedCoder);
|
|
this.badReferences = new HandleSet(WordReferenceRow.urlEntryRow.primaryKeyLength, WordReferenceRow.urlEntryRow.objectOrder, 0);
|
|
this.targets = targets;
|
|
this.hit = 0;
|
|
this.miss = 0;
|
|
}
|
|
|
|
/*
|
|
* return a new container with at most max elements and put the rest back to the index
|
|
* as this chunk might be transferred back to myself a random selection needs to be taken
|
|
* @param container
|
|
* @param max
|
|
* @throws RowSpaceExceededException
|
|
* @return
|
|
*/
|
|
private ReferenceContainer<WordReference> trimContainer(ReferenceContainer<WordReference> container, final int max) throws RowSpaceExceededException {
|
|
final ReferenceContainer<WordReference> c = new ReferenceContainer<WordReference>(Segment.wordReferenceFactory, container.getTermHash(), max);
|
|
final int part = container.size() / max + 1;
|
|
final Random r = new Random();
|
|
WordReference w;
|
|
List<byte[]> selected = new ArrayList<byte[]>();
|
|
final Iterator<WordReference> i = container.entries();
|
|
while ((i.hasNext()) && (c.size() < max)) {
|
|
w = i.next();
|
|
if (r.nextInt(part) == 0) {
|
|
c.add(w);
|
|
selected.add(w.urlhash());
|
|
}
|
|
}
|
|
// remove the selected entries from container
|
|
for (final byte[] b : selected) container.removeReference(b);
|
|
// put container back
|
|
try {
|
|
segment.termIndex().add(container);
|
|
} catch (Exception e) {
|
|
Log.logException(e);
|
|
}
|
|
return c;
|
|
}
|
|
|
|
/**
|
|
* add a container to the Entry cache.
|
|
* all entries in the container are checked and only such are stored which have a reference entry
|
|
* @param container
|
|
* @throws RowSpaceExceededException
|
|
*/
|
|
public void add(ReferenceContainer<WordReference> container) throws RowSpaceExceededException {
|
|
int remaining = maxRWIsCount;
|
|
for (ReferenceContainer ic : this) remaining -= ic.size();
|
|
if (remaining <= 0) {
|
|
// No space left in this chunk
|
|
try {
|
|
segment.termIndex().add(container);
|
|
} catch (Exception e) {
|
|
Log.logException(e);
|
|
}
|
|
return;
|
|
}
|
|
final ReferenceContainer<WordReference> c = (remaining >= container.size()) ? container : trimContainer(container, remaining);
|
|
// iterate through the entries in the container and check if the reference is in the repository
|
|
Iterator<WordReference> i = c.entries();
|
|
List<byte[]> notFoundx = new ArrayList<byte[]>();
|
|
while (i.hasNext()) {
|
|
WordReference e = i.next();
|
|
if (references.containsKey(e.urlhash())) continue;
|
|
if (badReferences.has(e.urlhash())) {
|
|
notFoundx.add(e.urlhash());
|
|
continue;
|
|
}
|
|
URIMetadataRow r = segment.urlMetadata().load(e.urlhash());
|
|
if (r == null) {
|
|
notFoundx.add(e.urlhash());
|
|
badReferences.put(e.urlhash());
|
|
} else {
|
|
references.put(e.urlhash(), r);
|
|
}
|
|
}
|
|
// now delete all references that were not found
|
|
for (final byte[] b : notFoundx) c.removeReference(b);
|
|
// finally add the remaining container to the cache
|
|
containers.add(c);
|
|
}
|
|
|
|
/**
|
|
* get all containers from the entry. This method may be used to flush remaining entries
|
|
* if they had been finished transmission without success (not enough peers arrived)
|
|
*/
|
|
public Iterator<ReferenceContainer<WordReference>> iterator() {
|
|
return this.containers.iterator();
|
|
}
|
|
|
|
public int containersSize() {
|
|
return this.containers.size();
|
|
}
|
|
|
|
public byte[] primaryTarget() {
|
|
return this.primaryTarget;
|
|
}
|
|
|
|
/**
|
|
* return the number of successful transmissions
|
|
* @return
|
|
*/
|
|
public int hit() {
|
|
return this.hit;
|
|
}
|
|
|
|
/**
|
|
* return the number of unsuccessful transmissions
|
|
* @return
|
|
*/
|
|
public int miss() {
|
|
return this.miss;
|
|
}
|
|
|
|
/**
|
|
* return the number of targets that are left in the target cache
|
|
* if this is empty, there may be no more use of this object and it should be flushed
|
|
* with the iterator method
|
|
* @return
|
|
*/
|
|
public int targets() {
|
|
return this.targets.size();
|
|
}
|
|
|
|
public boolean transmit() {
|
|
if (this.targets.isEmpty()) return false;
|
|
Seed target = this.targets.remove(0);
|
|
// transferring selected words to remote peer
|
|
if (target == seeds.mySeed() || target.hash.equals(seeds.mySeed().hash)) {
|
|
// target is my own peer. This is easy. Just restore the indexContainer
|
|
restore();
|
|
this.hit++;
|
|
log.logInfo("Transfer of chunk to myself-target");
|
|
return true;
|
|
}
|
|
log.logInfo("starting new index transmission request to " + ASCII.String(this.primaryTarget));
|
|
long start = System.currentTimeMillis();
|
|
final String error = Protocol.transferIndex(target, this.containers, this.references, gzipBody4Transfer, timeout4Transfer);
|
|
if (error == null) {
|
|
// words successfully transfered
|
|
long transferTime = System.currentTimeMillis() - start;
|
|
Iterator<ReferenceContainer<WordReference>> i = this.containers.iterator();
|
|
ReferenceContainer<WordReference> firstContainer = (i == null) ? null : i.next();
|
|
log.logInfo("Index transfer of " + this.containers.size() +
|
|
" words [" + ((firstContainer == null) ? null : ASCII.String(firstContainer.getTermHash())) + " .. " + ASCII.String(this.primaryTarget) + "]" +
|
|
" and " + this.references.size() + " URLs" +
|
|
" to peer " + target.getName() + ":" + target.hash +
|
|
" in " + (transferTime / 1000) +
|
|
" seconds successful (" + (1000 * this.containers.size() / (transferTime + 1)) +
|
|
" words/s)");
|
|
seeds.mySeed().incSI(this.containers.size());
|
|
seeds.mySeed().incSU(this.references.size());
|
|
// if the peer has set a pause time and we are in flush mode (index transfer)
|
|
// then we pause for a while now
|
|
log.logInfo("Transfer finished of chunk to target " + target.hash + "/" + target.getName());
|
|
this.hit++;
|
|
return true;
|
|
}
|
|
this.miss++;
|
|
// write information that peer does not receive index transmissions
|
|
log.logInfo("Transfer failed of chunk to target " + target.hash + "/" + target.getName() + ": " + error);
|
|
// get possibly newer target Info
|
|
Seed newTarget = seeds.get(target.hash);
|
|
if (newTarget != null) {
|
|
String oldAddress = target.getPublicAddress();
|
|
if ((oldAddress != null) && (oldAddress.equals(newTarget.getPublicAddress()))) {
|
|
newTarget.setFlagAcceptRemoteIndex(false);
|
|
seeds.update(newTarget.hash, newTarget);
|
|
} else {
|
|
// we tried an old Address. Don't change anything
|
|
}
|
|
} else {
|
|
// target not in DB anymore. ???
|
|
}
|
|
return false;
|
|
}
|
|
|
|
public boolean isFinished() {
|
|
//System.out.println("canFinish: hit = " + this.hit + ", redundancy = " + seeds.redundancy() + ", targets.size() = " + targets.size());
|
|
return this.hit >= seeds.redundancy();
|
|
}
|
|
|
|
public boolean canFinish() {
|
|
//System.out.println("canFinish: hit = " + this.hit + ", redundancy = " + seeds.redundancy() + ", targets.size() = " + targets.size());
|
|
return this.targets.size() >= seeds.redundancy() - this.hit;
|
|
}
|
|
|
|
public void restore() {
|
|
for (ReferenceContainer<WordReference> ic : this) try {
|
|
segment.termIndex().add(ic);
|
|
} catch (Exception e) {
|
|
Log.logException(e);
|
|
}
|
|
}
|
|
|
|
}
|
|
|
|
}
|