X-Git-Url: http://git.jankratochvil.net/?a=blobdiff_plain;f=project%2Ftcp_rto%2FListItem.pm;h=01a13ca06111028e10a4c7fff9173ef7edabc971;hb=0845bc10b5535a973faefa0ea76fc24af598af88;hp=7338b40f63acf6aaae7394433ac450027f1314de;hpb=4eff5fe31d7bf25f4471ff4329fde56cba97fdcf;p=www.jankratochvil.net.git diff --git a/project/tcp_rto/ListItem.pm b/project/tcp_rto/ListItem.pm index 7338b40..01a13ca 100755 --- a/project/tcp_rto/ListItem.pm +++ b/project/tcp_rto/ListItem.pm @@ -18,36 +18,40 @@ # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -package My::Project::tcp_rto; +package project::tcp_rto::ListItem; require 5.6.0; # at least 'use warnings;' but we need some 5.6.0+ modules anyway -use vars qw($VERSION $CVS_ID); -$VERSION=do { my @r=(q$Revision$=~/\d+/g); sprintf "%d.".("%03d"x$#r),@r; }; -$CVS_ID=q$Id$; +our $VERSION=do { my @r=(q$Revision$=~/\d+/g); sprintf "%d.".("%03d"x$#r),@r; }; +our $CVS_ID=q$Id$; use strict; use warnings; +use My::Web; -our %ListItem=( - "name"=>"Linux Kernel Patch for Maximal Round-trip-time", + +our @ListItem=( + "name"=>"tcp_rto", + "platform"=>"unixuser", + "trivia"=>1, "priority"=>7, "download-Linux kernel 2.2.17 patch"=>"linux-2.2.17-tcp_rto-1.diff", "download-Linux kernel 2.4.16 patch"=>"linux-2.4.16-tcp_rto-1.diff", - "summary"=>"Network workaround", + "summary"=>"Linux kernel maximal TCP round-trip-time patch", "license"=>"GPL", - "maintenance"=>"finished", + "maintenance"=>"ready", "language"=>"C patch", - "description"=>"" - .'

This patch can solve your problems if you have network connection dropping too much' - .' packets. In standard case the Linux kernel will correctly increase our round-trip-time' - .' of connection slowing the transfer rate up to the unusable state.

' - .'

After applying this patch you can set your maximal round-trip-time in file' - .' "/proc/sys/net/ipv4/tcp_rto_max", it is expressed in Hz' - .' units (100-per-second on x86 platform). You may need also to enlarge' - .' your maximal retry count in "/proc/sys/net/ipv4/tcp_retries2",' - .' otherwise your machine will reject the connection as it will have to retry' - .' the packets more than in sane states.

' - .'

Please use this feature very carefully! You are violating' - .' RFC standards and you can get your network administrators to be very angry!

' + "description"=><<"HERE", +

This patch can solve your problems if you have network connection dropping too much +packets. In standard case the Linux kernel will correctly increase our round-trip-time +of connection slowing the transfer rate up to the unusable state.

+

After applying this patch you can set your maximal round-trip-time in file +"/proc/sys/net/ipv4/tcp_rto_max", it is expressed in Hz +units (100-per-second on x86 platform). You may need also to enlarge +your maximal retry count in "/proc/sys/net/ipv4/tcp_retries2", +otherwise your machine will reject the connection as it will have to retry +the packets more than in sane states.

+

Please use this feature very carefully! You are violating +RFC standards and you can get your network administrators to be very angry!

+HERE ); 1;