introduction of dht-burst modes: this can expand the number of target peers in some cases where a better heuristic is needed. The problematic cases are either when a muti-word search is made (still a hard case for our term-oriented DHT) or when a network operator wants that all robinson peers are asked. We therefore introduced two new network steering values that switch on more peers during the peer selection. Because the number of peers can now be very large, the number of maximum httpc connections was also increased.

Please see new coments in yacy.network.freeworld.unit for details of the new DHT selection methods.
The number of maximum peers is now not fixed to a specific number but may increase with
- the partition exponent
- the number of redundant peers
- the robinson burst percentage
- the multiword burst percentage
The maximum can then be the number of senior peers (all visible peers).

git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7479 6c8d7289-2bf4-0310-a012-ef5d649a1542

1 files changed, 52 insertions, 1 deletions

diff --git a/defaults/yacy.network.freeworld.unit b/defaults/yacy.network.freeworld.unit
index 120370c7d..a484ec455 100644
--- a/defaults/yacy.network.freeworld.unit
+++ b/defaults/yacy.network.freeworld.unit
@@ -6,17 +6,68 @@
 # this is a work in progress. disabled properties are not yet used #
 # -----------------------------------------------------------------#
 
-# general network definition
+# define the name of the nework
+# this nickname is also used to identifiy network requests
 network.unit.name = freeworld
+
+# the visible name of the network
 network.unit.description = Public YaCy Community
+
+# definition of the content domain: possible values are:
+# global, local, any
 network.unit.domain = global
+
+# maximum search time for remote queries (deprecated)
 network.unit.search.time = 4
+
+# flag to switch on dht transmission
+# if the dht transmission is set to 'false' then for a global
+# query all targets are accessed
 network.unit.dht = true
+
+# the number of redundant target peers:
+# redundant peers get a copy of the original dht target information
 network.unit.dhtredundancy.junior = 1
 network.unit.dhtredundancy.senior = 3
+
+# the vertical partition of the dht: this applies a division
+# of the dht into 2^^<partitionExponent> fragments which get
+# all the same word-partition targets but a document-dht computed
+# fragment of all documents
 network.unit.dht.partitionExponent = 4
+
+# network request burst attributes: this enables non-dht target
+# positions for certain situations. This is not a 'traditional' burst-mode
+# since it does not refer to a handshake to a single client but it refers
+# to not-handshaking in a distributed way. It means to get data without using
+# a dht transmission logic.
+
+# robinson burst: percentage of the number of robinson peers that
+# shall be accessed for every search. This includes also robinson peers
+# that do not have a matching peer tag. If this is set to 100 then all robinson
+# peers are always asked
+network.unit.dht.burst.robinson = 50
+
+# multi-word burst: percentage of the number of all peers that
+# shall be accessed for multi-word searches. Multi-word search is
+# a hard problem when the distributed search network is divided by
+# term (as done with yacy, partly..).
+# Scientific solutions for this problem is to apply heuristics.
+# This heuristic enables to switch on a full network scan to get also
+# non-distributed multi-word positions. For a full scan set this value to 100.
+# Attention: this may out-number the maxcount of available httpc network connections.
+network.unit.dht.burst.multiword = 30
+
+# switch to enable verification of search results
+# must be set to true in untrusted networks and can be
+# set to false in completely trusted networks
 network.unit.inspection.searchverify = true
+
+# speed of remote crawl de-queueing. this is the number of milliseconds
+# as a pause between two requests
 network.unit.remotecrawl.speed = 300
+
+# addresses of seed-list bootstrap locations
 network.unit.bootstrap.seedlist0 = http://www.yacy.net/seed.txt
 network.unit.bootstrap.seedlist1 = http://home.arcor.de/hermens/yacy/seed.txt
 network.unit.bootstrap.seedlist2 = http://low.audioattack.de/yacy/seed.txt