- fixed more cluster routing problems

- fixed a problem in remote search when balancer caused shift process to wait too long

git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@3627 6c8d7289-2bf4-0310-a012-ef5d649a1542

htroot/ViewProfile.java

@@ -118,6 +118,7 @@ public class ViewProfile {
                 } catch (IOException e) {}
                 
                 // try to get the profile from remote peer
+                seed.setAlternativeAddress((String) switchboard.clusterhashes.get(seed.hash));
                 profile = yacyClient.getProfile(seed);
                 
                 // if profile did not arrive, say that peer is disconnected

htroot/yacy/hello.java

@@ -95,12 +95,14 @@ public final class hello {
         final String reportedPeerType = remoteSeed.get(yacySeed.PEERTYPE, yacySeed.PEERTYPE_JUNIOR);
         final float clientversion = remoteSeed.getVersion();
 
-        if ((sb.isRobinsonMode()) && (!sb.isOpenRobinsonCluster())) {
+        if ((sb.isRobinsonMode()) && (!sb.isPublicRobinson())) {
         	// if we are a robinson cluster, answer only if this client is known by our network definition
         	return prop;
         }
         
         int urls = -1;
+        remoteSeed.setAlternativeAddress((String) sb.clusterhashes.get(remoteSeed.hash));
+        
         // if the remote client has reported its own IP address and the client supports
         // the port forwarding feature (if client version >= 0.383) then we try to 
         // connect to the reported IP address first

htroot/yacy/message.java

@@ -97,7 +97,7 @@ public final class message {
         }
 
         if ((sb.isRobinsonMode()) &&
-        	 (!((sb.isOpenRobinsonCluster()) ||
+        	 (!((sb.isPublicRobinson()) ||
         	    (sb.isInMyCluster((String)header.get(httpHeader.CONNECTION_PROP_CLIENTIP)))))) {
             // if we are a robinson cluster, answer only if this client is known by our network definition
         	prop.putASIS("response", "-1"); // request rejected

htroot/yacy/profile.java

@@ -69,8 +69,8 @@ public final class profile {
         if (prop == null) { return null; }
 
         if ((sb.isRobinsonMode()) &&
-           	 (!((sb.isOpenRobinsonCluster()) ||
-           	    (sb.isInMyCluster((String)header.get(httpHeader.CONNECTION_PROP_CLIENTIP)))))) {
+           	(!sb.isPublicRobinson()) &&
+           	(!sb.isInMyCluster((String)header.get(httpHeader.CONNECTION_PROP_CLIENTIP)))) {
                // if we are a robinson cluster, answer only if this client is known by our network definition
         	prop.put("list", 0);
             return prop;

htroot/yacy/query.java

@@ -65,7 +65,7 @@ public final class query {
         if (prop == null || sb == null) { return null; }
 
         if ((sb.isRobinsonMode()) &&
-              	 (!((sb.isOpenRobinsonCluster()) ||
+              	 (!((sb.isPublicRobinson()) ||
               	    (sb.isInMyCluster((String)header.get(httpHeader.CONNECTION_PROP_CLIENTIP)))))) {
                   // if we are a robinson cluster, answer only if this client is known by our network definition
         	prop.putASIS("response", "-1"); // request rejected

htroot/yacy/search.java

@@ -117,7 +117,7 @@ public final class search {
         serverObjects prop = new serverObjects();
         
         if ((sb.isRobinsonMode()) &&
-             	 (!((sb.isOpenRobinsonCluster()) ||
+             	 (!((sb.isPublicRobinson()) ||
              	    (sb.isInMyCluster((String)header.get(httpHeader.CONNECTION_PROP_CLIENTIP)))))) {
                  // if we are a robinson cluster, answer only if this client is known by our network definition
         	prop.putASIS("links", "");

source/de/anomic/kelondro/kelondroCloneableMapIterator.java

@@ -1,4 +1,4 @@
-// kelondroCloneableSetIterator.java
+// kelondroCloneableMapIterator.java
 // (C) 2007 by Michael Peter Christen; mc@yacy.net, Frankfurt a. M., Germany
 // first published 25.04.2007 on http://yacy.net
 //
@@ -28,33 +28,33 @@
 package de.anomic.kelondro;
 
 import java.util.Iterator;
-import java.util.TreeSet;
+import java.util.TreeMap;
 
-public class kelondroCloneableSetIterator implements kelondroCloneableIterator {
+public class kelondroCloneableMapIterator implements kelondroCloneableIterator {
 
-	TreeSet set;
+	TreeMap map;
 	Object next, last;
 	Object start;
 	Iterator iter;
 	
 
-	public kelondroCloneableSetIterator(TreeSet set, Object start) {
+	public kelondroCloneableMapIterator(TreeMap map, Object start) {
 		// set must contain byte[] elements or String elements.
 		// start must be either of type byte[] or String
-		this.set = set;
+		this.map = map;
 		this.start = start;
-		this.iter = set.iterator();
+		this.iter = map.keySet().iterator();
 		if (this.start == null) {
 			if (iter.hasNext()) this.next = iter.next(); else this.next = null;
 		} else while (iter.hasNext()) {
 			this.next = iter.next();
-			if (set.comparator().compare(next, start) > 1) break;
+			if (map.comparator().compare(next, start) > 1) break;
 		}
 		this.last = null;
 	}
 	
 	public Object clone(Object modifier) {
-		return new kelondroCloneableSetIterator(set, modifier);
+		return new kelondroCloneableMapIterator(map, modifier);
 	}
 
 	public boolean hasNext() {
@@ -62,6 +62,7 @@ public class kelondroCloneableSetIterator implements kelondroCloneableIterator {
 	}
 
 	public Object next() {
+		// returns key-elements, not entry-elements
 		this.last = this.next;
 		if (this.iter.hasNext()) {
 			this.next = this.iter.next();
@@ -72,7 +73,7 @@ public class kelondroCloneableSetIterator implements kelondroCloneableIterator {
 	}
 
 	public void remove() {
-		this.set.remove(this.last);
+		this.map.remove(this.last);
 	}
 
 }

source/de/anomic/plasma/plasmaCrawlBalancer.java

@@ -397,6 +397,7 @@ public class plasmaCrawlBalancer {
         
         // finally: check minimumDelta and if necessary force a sleep
         long delta = lastAccessDelta(result);
+        assert delta >= 0: "delta = " + delta;
         if (delta < minimumDelta) {
             // force a busy waiting here
             // in best case, this should never happen if the balancer works propertly

source/de/anomic/plasma/plasmaSearchEvent.java

@@ -299,7 +299,7 @@ public final class plasmaSearchEvent extends Thread implements Runnable {
                 System.out.println("DEBUG-INDEXABSTRACT ***: peer " + peer + " from words: " + words);
                 secondarySearchThreads[c++] = yacySearch.secondaryRemoteSearch(
                         words, "", urls, urlStore, wordIndex, rcContainers, peer, plasmaSwitchboard.urlBlacklist, snippetCache,
-                        profileGlobal, ranking, query.constraint);
+                        profileGlobal, ranking, query.constraint, preselectedPeerHashes);
 
             }
         }

source/de/anomic/plasma/plasmaSwitchboard.java

@@ -1336,7 +1336,7 @@ public final class plasmaSwitchboard extends serverAbstractSwitch implements ser
     	return !getConfigBool(plasmaSwitchboard.INDEX_DIST_ALLOW, false) && !getConfigBool(plasmaSwitchboard.INDEX_RECEIVE_ALLOW, false);
     }
 
-    public boolean isOpenRobinsonCluster() {
+    public boolean isPublicRobinson() {
     	// robinson peers may be member of robinson clusters, which can be public or private
     	// this does not check the robinson attribute, only the specific subtype of the cluster
     	String clustermode = getConfig("cluster.mode", "publicpeer");
@@ -1999,24 +1999,20 @@ public final class plasmaSwitchboard extends serverAbstractSwitch implements ser
             //log.logDebug("LimitCrawl: queue is empty");
             return false;
         }
+        boolean robinsonPrivateCase = ((isRobinsonMode()) && 
+            	(!getConfig("cluster.mode", "").equals("publiccluster")) &&
+            	(!getConfig("cluster.mode", "").equals("privatecluster")));
         
-        if ((isRobinsonMode()) &&
-        	(!getConfig("cluster.mode", "").equals("publicpeer")) &&
-        	(!getConfig("cluster.mode", "").equals("privatepeer"))) { 
-        	// not-clustered robinson peers do not do remote crawling
-        	return false;
-        }
-        
-        if ((coreCrawlJobSize() <= 20) && (limitCrawlTriggerJobSize() > 100)) {
+        if ((robinsonPrivateCase) || ((coreCrawlJobSize() <= 20) && (limitCrawlTriggerJobSize() > 10))) {
             // it is not efficient if the core crawl job is empty and we have too much to do
             // move some tasks to the core crawl job
-            int toshift = limitCrawlTriggerJobSize() / 5;
-            if (toshift > 1000) toshift = 1000;
+            int toshift = 10; // this cannot be a big number because the balancer makes a forced waiting if it cannot balance
             if (toshift > limitCrawlTriggerJobSize()) toshift = limitCrawlTriggerJobSize();
             for (int i = 0; i < toshift; i++) {
                 noticeURL.shift(plasmaCrawlNURL.STACK_TYPE_LIMIT, plasmaCrawlNURL.STACK_TYPE_CORE);
             }
-            log.logInfo("shifted " + toshift + " jobs from global crawl to local crawl");
+            log.logInfo("shifted " + toshift + " jobs from global crawl to local crawl (coreCrawlJobSize()=" + coreCrawlJobSize() + ", limitCrawlTriggerJobSize()=" + limitCrawlTriggerJobSize() + ", cluster.mode=" + getConfig("cluster.mode", "") + ", robinsonMode=" + ((isRobinsonMode()) ? "on" : "off"));
+            if (robinsonPrivateCase) return false;
         }
         
         // if the server is busy, we do crawling more slowly
@@ -2484,6 +2480,7 @@ public final class plasmaSwitchboard extends serverAbstractSwitch implements ser
                         // if this was performed for a remote crawl request, notify requester
                         if ((processCase == PROCESSCASE_6_GLOBAL_CRAWLING) && (initiatorPeer != null)) {
                             log.logInfo("Sending crawl receipt for '" + entry.normalizedURLString() + "' to " + initiatorPeer.getName());
+                            initiatorPeer.setAlternativeAddress((String) clusterhashes.get(initiatorPeer.hash));
                             yacyClient.crawlReceipt(initiatorPeer, "crawl", "fill", "indexed", newEntry, "");
                         }
                     } else {
@@ -2498,6 +2495,7 @@ public final class plasmaSwitchboard extends serverAbstractSwitch implements ser
                     
                     log.logSevere("Could not index URL " + entry.url() + ": " + ee.getMessage(), ee);
                     if ((processCase == PROCESSCASE_6_GLOBAL_CRAWLING) && (initiatorPeer != null)) {
+                    	initiatorPeer.setAlternativeAddress((String) clusterhashes.get(initiatorPeer.hash));
                         yacyClient.crawlReceipt(initiatorPeer, "crawl", "exception", ee.getMessage(), null, "");
                     }
                     addURLtoErrorDB(entry.url(), referrerUrlHash, initiatorPeerHash, docDescription, plasmaCrawlEURL.DENIED_UNSPECIFIED_INDEXING_ERROR, new kelondroBitfield());
@@ -2510,6 +2508,7 @@ public final class plasmaSwitchboard extends serverAbstractSwitch implements ser
                 log.logInfo("Not indexed any word in URL " + entry.url() + "; cause: " + noIndexReason);
                 addURLtoErrorDB(entry.url(), referrerUrlHash, initiatorPeerHash, docDescription, noIndexReason, new kelondroBitfield());
                 if ((processCase == PROCESSCASE_6_GLOBAL_CRAWLING) && (initiatorPeer != null)) {
+                	initiatorPeer.setAlternativeAddress((String) clusterhashes.get(initiatorPeer.hash));
                     yacyClient.crawlReceipt(initiatorPeer, "crawl", "rejected", noIndexReason, null, "");
                 }
             }
@@ -2670,7 +2669,7 @@ public final class plasmaSwitchboard extends serverAbstractSwitch implements ser
         }
         
         // check if peer for remote crawl is available
-        yacySeed remoteSeed = ((this.isOpenRobinsonCluster()) && (getConfig("cluster.mode", "").equals("publiccluster"))) ?
+        yacySeed remoteSeed = ((this.isPublicRobinson()) && (getConfig("cluster.mode", "").equals("publiccluster"))) ?
         	yacyCore.dhtAgent.getPublicClusterCrawlSeed(urlEntry.urlhash(), this.clusterhashes) :	
         	yacyCore.dhtAgent.getGlobalCrawlSeed(urlEntry.urlhash());
         if (remoteSeed == null) {
@@ -2694,6 +2693,7 @@ public final class plasmaSwitchboard extends serverAbstractSwitch implements ser
             return false; // no response from peer, we will crawl this ourself
         }
         
+        String response = (String) page.get("response");
         log.logFine("plasmaSwitchboard.processRemoteCrawlTrigger: remoteSeed="
                 + remoteSeed.getName() + ", url=" + urlEntry.url().toString()
                 + ", response=" + page.toString()); // DEBUG
@@ -2701,7 +2701,6 @@ public final class plasmaSwitchboard extends serverAbstractSwitch implements ser
         // we received an answer and we are told to wait a specific time until we shall ask again for another crawl
         int newdelay = Integer.parseInt((String) page.get("delay"));
         yacyCore.dhtAgent.setCrawlDelay(remoteSeed.hash, newdelay);
-        String response = (String) page.get("response");
         if (response.equals("stacked")) {
             // success, the remote peer accepted the crawl
             log.logInfo(STR_REMOTECRAWLTRIGGER + remoteSeed.getName()

source/de/anomic/yacy/yacyCore.java

@@ -349,10 +349,10 @@ public class yacyCore {
                                       seedDB.sizeConnected() + " different peers");
     }
 
-    private boolean canReachMyself() {
+    private boolean canReachMyself() { // TODO: check if this method is necessary - depending on the used router it will not work
         // returns true if we can reach ourself under our known peer address
         // if we cannot reach ourself, we call a forced publishMySeed and return false
-        final int urlc = yacyClient.queryUrlCount(seedDB.mySeed);
+    	final int urlc = yacyClient.queryUrlCount(seedDB.mySeed);
         if (urlc >= 0) {
             seedDB.mySeed.setLastSeenUTC();
             return true;

source/de/anomic/yacy/yacyDHTAction.java

@@ -47,11 +47,10 @@ import java.util.ArrayList;
 import java.util.Enumeration;
 import java.util.Hashtable;
 import java.util.TreeMap;
-import java.util.TreeSet;
 
 import de.anomic.kelondro.kelondroBase64Order;
 import de.anomic.kelondro.kelondroCloneableIterator;
-import de.anomic.kelondro.kelondroCloneableSetIterator;
+import de.anomic.kelondro.kelondroCloneableMapIterator;
 import de.anomic.kelondro.kelondroException;
 import de.anomic.kelondro.kelondroMScoreCluster;
 import de.anomic.kelondro.kelondroRotateIterator;
@@ -219,9 +218,10 @@ public class yacyDHTAction implements yacyPeerAction {
     }
     
     public synchronized yacySeed getPublicClusterCrawlSeed(String urlHash, TreeMap clusterhashes) {
-        kelondroCloneableIterator i = new kelondroRotateIterator(new kelondroCloneableSetIterator((TreeSet) (clusterhashes.keySet()), urlHash), null);
-        if (i.hasNext()) {
+        kelondroCloneableIterator i = new kelondroRotateIterator(new kelondroCloneableMapIterator(clusterhashes, urlHash), null);
+        while (i.hasNext()) {
         	yacySeed seed = seedDB.getConnected((String) i.next());
+        	if (seed == null) continue;
         	seed.setAlternativeAddress((String) clusterhashes.get(seed.hash));
         	return seed;
         }

source/de/anomic/yacy/yacySearch.java

@@ -271,13 +271,14 @@ public class yacySearch extends Thread {
             indexContainer containerCache,
             String targethash, plasmaURLPattern blacklist, plasmaSnippetCache snippetCache,
             plasmaSearchTimingProfile timingProfile, plasmaSearchRankingProfile rankingProfile,
-            kelondroBitfield constraint) {
+            kelondroBitfield constraint, TreeMap clusterselection) {
         // check own peer status
         if (yacyCore.seedDB.mySeed == null || yacyCore.seedDB.mySeed.getPublicAddress() == null) { return null; }
 
         // prepare seed targets and threads
         final yacySeed targetPeer = yacyCore.seedDB.getConnected(targethash);
         if (targetPeer == null) return null;
+        targetPeer.setAlternativeAddress((String) clusterselection.get(targetPeer.hash));
         yacySearch searchThread = new yacySearch(wordhashes, excludehashes, urlhashes, "", "", 9999, true, 0, targetPeer,
                                              urlManager, wordIndex, containerCache, new TreeMap(), blacklist, snippetCache, timingProfile, rankingProfile, constraint);
         searchThread.start();

source/de/anomic/yacy/yacySeedDB.java

@@ -59,19 +59,15 @@ import java.util.Hashtable;
 import java.util.Iterator;
 import java.util.Map;
 import java.util.TreeMap;
-import java.util.TreeSet;
 
 import de.anomic.http.httpHeader;
 import de.anomic.http.httpc;
 import de.anomic.http.httpd;
-import de.anomic.kelondro.kelondroCloneableIterator;
-import de.anomic.kelondro.kelondroCloneableSetIterator;
 import de.anomic.kelondro.kelondroDyn;
 import de.anomic.kelondro.kelondroException;
 import de.anomic.kelondro.kelondroMScoreCluster;
 import de.anomic.kelondro.kelondroMapObjects;
 import de.anomic.kelondro.kelondroBase64Order;
-import de.anomic.kelondro.kelondroRotateIterator;
 import de.anomic.net.URL;
 import de.anomic.plasma.plasmaSwitchboard;
 import de.anomic.server.serverCore;
@@ -296,21 +292,6 @@ public final class yacySeedDB {
     	return clustermap;
     }
     
-    public Iterator /*of yacySeed*/ seedsInCluster(String firstHash, TreeSet clusterhashes) {
-    	// returns an seed iterator for all peer hashes in the given set
-    	// the iterator starts at the firstHash
-        kelondroCloneableIterator i = new kelondroRotateIterator(new kelondroCloneableSetIterator(clusterhashes, firstHash), null);
-        ArrayList l = new ArrayList();
-        Object o;
-        while (i.hasNext()) {
-        	o = i.next();
-        	if (o instanceof String) l.add(get((String) o));
-        	if (o instanceof byte[]) l.add(get(new String((byte[]) o)));
-        	if (l.size() >= clusterhashes.size()) break;
-        }
-        return l.iterator();
-    }
-    
     public Enumeration seedsConnected(boolean up, boolean rot, String firstHash, float minVersion) {
         // enumerates seed-type objects: all seeds sequentially without order
         return new seedEnum(up, rot, (firstHash == null) ? null : firstHash.getBytes(), null, seedActiveDB, minVersion);