Replication + load balancing

[stefan]

1. Replication of the website through litespeed.

Imagine for a company who has a backup-machine with litespeed running as hot-standby or just as backup. This company has to keep both machines up-to-date, maybe this can be handled by litespeed through replication. It is maybe far fetched, but it is a nice idea if you make modifications to files (webpages, css, php and so on) on the master webserver and this webserver will update the replication servers (backup , hotstandby ones) Or initiated by litespeed, through the admin interface, you can even use default unix commands!?


2. Load-balancing.

You can have load-balancing implemented if you have replication.
the backup/replication server can be a just read (of course) server. Maybe you can even configure in the admin interface how the load is balanced, say 80% to 20%, or whatever you like to have it! This can depend on how powerfull the other machine is.

[mistwang]

Thank you for those feature requests.
Load balancing is on our todo list.
Replication is a cool feature, it depends on our schedule.

However, you may not be able to see those features in standard edition. :-)

[xing]

Replication can be easily done with today's version software.

Let's say, you have /mylsws folder which contain all your lsws binary, config, and php/html files, you can just do this:

Subversion master server : /mylsws

primary webserver : run code update from master on /mylsws every 60 seconds

secondary webserver: run code update from master.....

For svn, just have to run "svn update" every 60 in cron and you have a full backup ready to rock-n-roll. I do this for my cluster slaves so that all them will have up to date code within 60 seconds when I change the code on the cvs server.

Replication of code/files is such a beast I say leave it to the specialty versioning softwares. =)

[stefan]

Originally Posted by xing

Replication can be easily done with today's version software.

Let's say, you have /mylsws folder which contain all your lsws binary, config, and php/html files, you can just do this:

Subversion master server : /mylsws

primary webserver : run code update from master on /mylsws every 60 seconds

secondary webserver: run code update from master.....

For svn, just have to run "svn update" every 60 in cron and you have a full backup ready to rock-n-roll. I do this for my cluster slaves so that all them will have up to date code within 60 seconds when I change the code on the cvs server.

Replication of code/files is such a beast I say leave it to the specialty versioning softwares. =)

Thanx, I will try it and report back.