Updated documentation for choke groups.

This commit is contained in:
rakshasa
2012-04-16 17:06:53 +09:00
parent 17201984a6
commit 63cf700593
4 changed files with 140 additions and 1 deletions
+25
View File
@@ -0,0 +1,25 @@
# Choke Groups
## Settings
choke_group.insert = "leech_fast"
Create a new group named "leech_fast", accessible by the string or
index / reverse index according to order of insertion. E.g. '-1'
refers to the last inserted choke group, while 0 refers to the first.
choke_group.tracker.mode.set = -1,"aggressive"
Set the tracker mode for torrents in this group.
choke_group.up.heuristics.set = -1,"upload_leech_experimental"
choke_group.down.heuristics.set = -1,"download_leech"
Set the heuristics used when deciding on which peers to choke and
unchoke. Use "strings.choke_heuristics{,.upload,.download}" to get a
list of the available heuristics.
choke_group.up.max.set = -1,250
choke_group.down.max.set = -1,500
Set the max total number of unchoked peers for all torrents in this choke group.
+64
View File
@@ -0,0 +1,64 @@
# IP filtering
## Introduction
## Make a new ip table
ip_tables.insert_table = <table_name>
Create a new empty table with the name table\_name, with the default
value returned being $0$.
There is currently no use of the generic ip tables commands.
## Add a new address block
ip_tables.add_address = <table_name>, 10.0.0.0/8, <value>
Set <value\> for all addresses in the address block, overwriting prior
values.
## Add a new address block
ip_tables.load = <table_name>, ~/foo.txt, <value>
Set <value\> for all addresses in the file foo.txt separated by
newline, similar to add\_address.
## Get value for address
ip_tables.get = <table_name>, 10.10.10.10
Returns the value set for an address, or the address block it belongs
to. The default is $0$.
## Size of data structures
ip_tables.size_data = <table_name>
Returns the size in bytes of all data structures for this table,
excluding the root class object itself. Note that the in-memory table is
dynamically consolidated, as such memory use will always be based on
actual fragmentation.
The table is a b-tree with 1024 nodes per branch.
## IPv4 filtering table
ipv4_filter.add_address = 10.0.0.0/8, unwanted
ipv4_filter.add_address = 11.0.0.0/8, preferred
ipv4_filter.load = ~/filters.txt, unwanted
ipv4_filter.get = 10.10.10.10
ipv4_filter.size_data =
The main ip filter, currently supporting unwanted (do not allow
connections) and preferred (currently used only in private code).
## Constants
strings.ip_filter =
=>
{ "unwanted", PeerInfo::flag_unwanted },
{ "preferred", PeerInfo::flag_preferred },
Constants used by ipv4\_filter values.
+1 -1
View File
@@ -1,4 +1,4 @@
\section(IP filtering}
\section{IP filtering}
\subsection{Introduction}
+50
View File
@@ -0,0 +1,50 @@
# Logging
## Opening log files
# log.open_file = "log name", "file path"
log.open_file = "rtorrent.log", (cat,/tmp/rtorrent.log.,(system.pid))
A newly opened log file is not connected to any logging events.
Some control over formatting will be provided at a later date.
## Adding outputs to events
# log.add_output = "logging event", "log name"
log.add_output = "info", "rtorrent.log"
log.add_output = "dht_debug", "tracker.log"
log.add_output = "tracker_debug", "tracker.log"
Each log handle can be added to multiple different logging events.
## Logging events
"critical"
"error"
"warn"
"notice"
"info"
"debug"
The above events receive logging events from all the sub-groups
displayed below, and each event also reciving events from the event
above in importance.
Thus some high-volume sub-group events such as “tracker\_debug” are not
part of “debug” and every “warn” event will receive events from “error”,
“critical”.
"connection_*"
"dht_*"
"peer_*"
"rpc_*"
"storage_*"
"thread_*"
"tracker_*"
"torrent_*"
All sub-groups have events from “critical” to “debug” defined.