On Wed, Oct 08, 2003 at 06:53:40PM +0800, Joon wrote: > I also think this aspect of the system will be the bulk of my problem > come implementation time. The servers will provide http and maybe > https services with a postgresql database-- the usual content management > stuff. I expect the developers to be modifying the code often, so it's > essential to sync the code files and databases (I don't have to sync the > logfiles in realtime, do I?). Although it would help if server B would > update binaries and config files too everytime I update or recompile > from server A. >
Without a third server, this can be difficult. You'll need a clustering database that does replication, simply copying data files onto a running database usually spells disaster. A number of Free software solutions exist (e.g. C-JDBC if you use Java, MySQL's and PostGreSQL's replication, and so on), if you can't shell out the resources for a third server to act as a NAS or money for a storage area network setup (these can cost millions, I kid you not). > Anyway, I guess the syncing thing is incremental. I'm now looking at > CODA and rsync. I'm not sure yet if these are what I really need. > CODA is known to be unstable. I wouldn't use it in a production environment just yet. > Ah yes, I'm looking at it now... I think this is enough! I also > stumbled upon FAKE which does the IP add takeover. Or is it that > Heartbeat natively uses FAKE? Heartbeat does essentially the same thing as Fake. I've only used Heartbeat (in an enterprise server setup no less), and can say that it's pretty solid. -- Philippine Linux Users' Group (PLUG) Mailing List [EMAIL PROTECTED] (#PLUG @ irc.free.net.ph) Official Website: http://plug.linux.org.ph Searchable Archives: http://marc.free.net.ph . To leave, go to http://lists.q-linux.com/mailman/listinfo/plug . Are you a Linux newbie? To join the newbie list, go to http://lists.q-linux.com/mailman/listinfo/ph-linux-newbie
