[Mediawiki-l] keeping my mirror in sync

[Platonides]

Waleed Meligy wrote:
> Thanks for your help. one more question:
> 
> We have started importing the wikipedia dump using ImportDump.php, the issue
> is we are using PuTTY and an SSH session to do that. We have kept a PC and
> the session open for 30+ hours now, we have currently imported 3,637,700
> pages.


a) You can make a program like
int main() {
if (fork()) return 0; //Parent
system("php importdump.php....."); //The command you use
return 0;
}

compile with gcc and run. The command will keep running even if you 
close the ssh client (or the network drops the connection).
(There must be an easier way, any taker between the unix gurus in the list?)


> How long do we have to wait ? and i am 80% sure there has to be a quicker
> way. Is there any ?
English wikipedia is loong to import. But using importDump is always 
much slower because it not only puts the data on the db, but also parses 
the page.
You should use mwdumper http://meta.wikimedia.org/wiki/Data_dumps#mwdumper