Hi Bong, Your question reminded me of my projects and PLUG presentations[1] back in 2002 -2005. I'm just surprised to hear that people are still recommending the use of Heartbeat/DRBD for implementing a redundant website infrastructure. Back then, DRBD isn't fast enough for our application and made half of our expensive servers operate in hot-standby mode (a.k.a. Failover Cluster). Keep in mind that operating the machine in hot standby mode doesn't scale when the website's traffic increases and even worse, it won't necessarily extend the life span of the machine (a common misconception).
Implementing a Failover cluster to run a redundant website infra sounds lazy to me. You should consider implementing a Load-Balanced Cluster instead. Start by looking into the following tools: - Load Balancer: BigIP or simply use Apache HTTPD's mod_proxy_balancer[2] - Replication of Web Applications: use a deployment tool[3][4] since you don't really need to replicate the applications in real time - Replication of Content Uploads: GlusterFS[5], etc. - Replication of MySQL: MySQL Replication or Cluster[6] - Handling of Session Data in a LB cluster: leverage the above-mentioned tools (e.g. LB's sticky sessions) or simply use Zend Server's Session Clustering[7] Ref: [1] https://docs.google.com/file/d/0By_VSgVKufmNNU9tclRZTG1iZmM [2] http://httpd.apache.org/docs/2.4/mod/mod_proxy_balancer.html [3] http://jenkins-ci.org/ [4] http://jenkins-php.org/ [5] http://www.gluster.org/ [6] http://dev.mysql.com/doc/refman/5.6/en/ha-overview.html [7] http://static.zend.com/topics/Session-Clustering-ZSCM-WP-0610-EN-A4.pdf _________________________________________________ Philippine Linux Users' Group (PLUG) Mailing List http://lists.linux.org.ph/mailman/listinfo/plug Searchable Archives: http://archives.free.net.ph
