- new limit to use the table copy (instead of flag): 400MB available. If
less is available, then a copy is never used. If more is available, then
it can be used if there is a remaining space of at least 200MB
- flush caches more often: flush the Digest cache
not equal. It might happen that these data is different because one of
that caches is cleaned after a while or when it is too big. The metadata
is then not cleaned, but now wiped after a checkup process at every
application start. This should cause a bit less memory usage.
- added log warnings in case that search processes run into time-out
situations
- better concurrency for Integer formatter (used a non-synchronized
formatter before)
- bugfix for search termination (a poison pill was missing)
- added timeout parameters for search (again) -> target is, that they
are never reached.
- search request are now made using a map boundary
- search results are only computed for the map boundary
- the number of results is adopted to the results in the visible range
- added a double-buffering for the search result markers
- added a search query option for the search results:
/radius/<lat>/<lon>/<radius>
SeedDB.lookupByName searche for lowercase peerNames, while MapColumnIndex.getIndex uses peername as is in the keyset.
Changed the index init to insert lowercase peer names as key
- a new database for url reference data ('seen links')
- a new database extending the references to the full url metadata
attributes set which shall replace the old metadata database if it is
finished
- migration help classes stub to use old and new metadata databases
simultanously
See
http://docs.oracle.com/javase/6/docs/api/java/util/concurrent/ConcurrentLinkedQueue.html
and the following test programm:
public class QueueLengthTimeTest {
public static long countTest(Queue<Integer> q, int c) {
long t = System.currentTimeMillis();
for (int i = 0; i < c; i++) {
q.add(q.size());
}
return System.currentTimeMillis() - t;
}
public static void main(String[] args) {
int c = 1;
for (int i = 0; i < 100; i++) {
Runtime.getRuntime().gc();
long t1 = countTest(new ArrayBlockingQueue<Integer>(c), c);
Runtime.getRuntime().gc();
long t2 = countTest(new LinkedBlockingQueue<Integer>(), c);
Runtime.getRuntime().gc();
long t3 = countTest(new ConcurrentLinkedQueue<Integer>(),
c);
System.out.println("count = " + c + ": ArrayBlockingQueue =
" + t1 + ", LinkedBlockingQueue = " + t2 + ", ConcurrentLinkedQueue = "
+ t3);
c = c * 2;
}
}
}
during receiving of DHT submissions and when answering remote search
requests. Both events together may have caused IO-deadlocking and this
commit shall fix that.
method. This is much faster and produces less blocking when using the
Compressor class which is used by the HTCache. I.e. picture search is
much faster now.
Hashtable is an obsolete collection v1, now since v2 offers HashMap with same or better
functionality. Please review, almost all code was already moved, so only a few changes. That is not the issue,
but I found notices that some (ugly big) helper classes had to be created in past
to compensate missing Hashtable's functionality. I'd like input if we can remove some of them.
look for //FIX: if these commits
Signed-off-by: Marek Otahal <markotahal@gmail.com>
http://forum.yacy-websuche.de/viewtopic.php?f=18&t=3461
The search can now be configured with a non-display host list.
the search will always exlude the given list of host unless they are
requested directly using the host navigation
better-ranked search results. without that pauses the result page will
only contain links from the peer that answers first which is not a good
average picture of all the peers that provided results
- using configuration-values of crawling-max-filesize also for snippetfetching and loading files into Index
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7893 6c8d7289-2bf4-0310-a012-ef5d649a1542
You can toggle between previous (standard) and new (generation) strategy at PerformanceMemory_p.html.
The generation memory strategy is implemented with the objective of running more robust
but with the cost of early stopping some tasks (eg. dht) while running low on memory.
This new strategy does respect the generational way a heap is organized on most used jvms.
These changes run fine on my 3 peers for weeks now, but as I'm human, I may fail.
Please be carefull using generation memory strategy and report errors by naming
OS, jvm and java_args.
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7886 6c8d7289-2bf4-0310-a012-ef5d649a1542
- added a button in PerformanceMemory_p.html to set the simulated short memory status
- bugfix: added a missing lowercase in KeyList
- better concurrency in loader dispatcher
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7883 6c8d7289-2bf4-0310-a012-ef5d649a1542
I wondered getting Total-ram > Max-ram and MemoryControl.available() < 0
MemoryControl.available() < 0 causes some errors where its value is used for dimension of buffers for eg.
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7852 6c8d7289-2bf4-0310-a012-ef5d649a1542
- added new Array class that contains an abstraction of the java Arrrays class which replaces the home-brew quicksort algorithm.
- the new class is about four times slower than the old one, but it works correct (the old one had errors)
- fixed a synchronization problem
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7842 6c8d7289-2bf4-0310-a012-ef5d649a1542
java.lang.NullPointerException
at net.yacy.kelondro.index.RowCollection.<init>(RowCollection.java:97)
at net.yacy.kelondro.index.RowSet.<init>(RowSet.java:48)
at net.yacy.kelondro.rwi.ReferenceContainer.<init>(ReferenceContainer.java:58)
at net.yacy.kelondro.rwi.ReferenceIterator.next(ReferenceIterator.java:69)
at net.yacy.kelondro.rwi.ReferenceIterator.next(ReferenceIterator.java:43)
at net.yacy.kelondro.blob.ArrayStack.merge(ArrayStack.java:1023)
at net.yacy.kelondro.blob.ArrayStack.mergeWorker(ArrayStack.java:922)
at net.yacy.kelondro.blob.ArrayStack.mergeMount(ArrayStack.java:869)
at net.yacy.kelondro.rwi.IODispatcher$MergeJob.merge(IODispatcher.java:267)
at net.yacy.kelondro.rwi.IODispatcher$MergeJob.access$300(IODispatcher.java:239)
at net.yacy.kelondro.rwi.IODispatcher.run(IODispatcher.java:180)
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7822 6c8d7289-2bf4-0310-a012-ef5d649a1542
for example search for "passwd /ftp". This can also be done with /http /https and /smb
- fixed some search throttling processes that should protect your peer against search DoS or strong search load
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7794 6c8d7289-2bf4-0310-a012-ef5d649a1542
- added a memory limitation in the zip parser and the pdf parser
- added a search throttling: if there are too many search queries are still to be computed, then new requests are not accepted for some time. if after a one second still no space is there to perform another search, the search terminates with no results. this case should only happen in case of DoS-like situations and in case of strong load on a peer like if it is integrated in metager.
- added a search cache deletion process that removes search requests in case that throttling happens
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7766 6c8d7289-2bf4-0310-a012-ef5d649a1542
- forced a possible short memory status when a search is started to flush caches that may cause search-heaps with resource contention effects
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7747 6c8d7289-2bf4-0310-a012-ef5d649a1542
used a ASCII String <-> byte[] conversion wherever possible. Many Strings in YaCy are hashes which are pure ASCII (base64 hashes).
The new ASCII String <-> byte[] conversion method have less computation overhead than the UTF8 conversion.
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7746 6c8d7289-2bf4-0310-a012-ef5d649a1542
- many speed/performance hacks
- added solr charding and new charding web interface
- added option to switch off the yacy index when using solr
- added new fail-url categories which are used to make a distinction which fail-urls to be sent to solr
- refactoring/renaming of some method names to distinguish host/url hashes better
- a large number of bug/npe fixes
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7738 6c8d7289-2bf4-0310-a012-ef5d649a1542
- less threads for blocking threads
- disable all threads for DHT transmission for networks with zero peers
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7737 6c8d7289-2bf4-0310-a012-ef5d649a1542
This time it works like this:
- each peer provides its ranking information using the yacy/idx.json servlet
- peers with more than 1 GB ram will load this information from all other peers, combine that into one ranking table and store it locally. This happens during the start-up of the peer concurrently. The new generated file with the ranking information is at DATA/INDEX/<network>/QUEUES/hostIndex.blob
- this index is then computed to generate a new fresh ranking table. Peers which can calculate their own ranking table will do that every start-up to get latest feature updates until the feature is stable
- I computed new ranking tables as part of the distribition and commit it here also
- the YBR feature must be enabled manually by setting the YBR value in the ranking servlet to level 15. A default configuration for that is also in the commit but it does not affect your current installation only fresh peers
- a recursive block rank refinement is implemented but disabled at this point. it needs more testing
Please play around with the ranking settings and see if this helped to make search results better.
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7729 6c8d7289-2bf4-0310-a012-ef5d649a1542
This servlet currently only serves for indexes to the web structure hosts. It can be tested by calling
http://localhost:8090/yacy/idx.json?object=host
This yacy protocol servlet is the first one that returns JSON code and that also shows index entries in a readable format. This will make the development of API applications much easier. This is also an example implementation for possible json versions of the other existing YaCy protocol interfaces.
The main purpose of this new feature is to provide a distributed block rank collection feature. Creating a block rank is very difficult if the forward-link data is first collected and then one peer must create a backward-link index. This interface provides already a partial backward index and therefore a collection of all these indexes needs only to be joined which is very easy. The result should be the computation of new block rank tables that all peers can perform.
To reduce load from peers this servlet buffers all data and refreshes it only once in 12 hours. This very slow update cycle is needed because the interface will be called round-robin from all peers once after start-up.
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7724 6c8d7289-2bf4-0310-a012-ef5d649a1542
- added transport of string-search pattern to remote search protocol
- fixed a problem parsing snippets with a '-' inside
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7700 6c8d7289-2bf4-0310-a012-ef5d649a1542
- added more properties to solr index
- refactoring
- more constants in switchboard
- fix for some NPEs
- recognition of more images
- removed synchronization in HandleMap (obviously not necessary?)
- added a nolocal configuration to remove excessive dns lookup (works only on allip - default off). Indexes produced with this setting are all flagged with 'local' and are (on purpose) not usable for freeworld because they will be rejected as beeing local.
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7672 6c8d7289-2bf4-0310-a012-ef5d649a1542
- added parser for in-text appearing geo-locations
- added geo-locations to rss search result
- added evaluation of metadata-attached geo-locations in yacysearch_location to show search results within a map
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7631 6c8d7289-2bf4-0310-a012-ef5d649a1542
Michael Peter Christen