different search results using this same index

Hier finden YaCy User Hilfe wenn was nicht funktioniert oder anders funktioniert als man dachte. Bei offensichtlichen Fehlern diese bitte gleich in die Bugs (http://bugs.yacy.net) eintragen.
Forumsregeln
In diesem Forum geht es um Benutzungsprobleme und Anfragen für Hilfe. Wird dabei ein Bug identifiziert, wird der thread zur Bearbeitung in die Bug-Sektion verschoben. Wer hier also einen Thread eingestellt hat und ihn vermisst, wird ihn sicherlich in der Bug-Sektion wiederfinden.

different search results using this same index

Beitragvon streetfighter » Sa Jan 03, 2009 10:00 am

I have independent p2p network (2 nodes) using yacy 0.617/05425 on java 1.6
With paused crawling and small delay I enter this same search word on both search pages. I think results should be this same but was different
1st node
1-10 results from a total number of 2,388 known (1,517 local, 871 remote), 10 links from 1 other YaCy peers
2nd node
1-10 results from a total number of 2,388 known (871 local, 1,517 remote), 10 links from 1 other YaCy peers

Can anybody tell me is it normal (why) or is it bug?
streetfighter
 
Beiträge: 37
Registriert: Sa Jan 03, 2009 9:40 am

Re: different search results using this same index

Beitragvon Orbiter » Mo Jan 05, 2009 2:56 pm

wow, an independent network! where did you learn how to do it? I believe the only documentation for that is in german, here:
http://www.yacy-websuche.de/wiki/index. ... definition

ok, two peers, different results. This is correct, because of timing. When a global search is started, the local search happens during waiting for the results from the remote search. In this time, possible result URLs are fetched from the URL database, and then passed to the 'secondary ranking', which uses rules that can only be done when the clear text of a URL is known. This fetching stops when remote search results appear, and then the results are combined. Different peers may have different timing, CPU load, different number of URLs in the local database and so on. All these things influence timing, which means that results from different peers are different. But because the Peers learn from each other, the results should be the same after a while.
You can try a test: open the Ranking servlet; don't change anything there, this is only to cause a search cache reset (which is done for different causes, one is the opening of that page). Then try your search again; results should now be more similar, I hope...
Orbiter
 
Beiträge: 5798
Registriert: Di Jun 26, 2007 10:58 pm
Wohnort: Frankfurt am Main

Re: different search results using this same index

Beitragvon streetfighter » Mi Jan 07, 2009 10:04 am

lang was not problem :) and google translate give me some fun :) with interpretation
I was fightning with this topic over 2 weeks (with this link 2 days)

There is no info about superseed.txt file but networks is working on seed.txt too

I will make new post in Wunschliste - network definition is cleared every update process this same with changes in html files, etc.
streetfighter
 
Beiträge: 37
Registriert: Sa Jan 03, 2009 9:40 am


Zurück zu Fragen und Antworten

Wer ist online?

Mitglieder in diesem Forum: 0 Mitglieder und 1 Gast

cron