Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On 26/02/13 20:59, Simone Del Pinto wrote: > Hi Lars, > > yes now is more clar for me. > > We made "node standby" to simulate a package switching ( in a wrong > way it seems ). > Let me ask one thing: do ypu know a way to do a "cluster functionalty" > test? > > I'm thinking about unplugging lan cable... is that a right test? > > Thanks again for your support. > I'm pretty sure it was netflix recently released a tool which will randomly kill amazon EC2 instances to simulate "failure situations". Basically, you want ANY, RANDOM item to fail, that could be a disk, a LAN cable, a switch, a complete machine. Walk into the server room, and look around, write down a list of everything you own, every little thing as well as the big things. Now work down the list, and decide what might happen if this item failed, if you think you should be protected, test it (not during prime time, something always unexpected happens). If you know you are not protected, decide what you need to do to protect it, then decide if that is worth it (chance of failure, ost of downtime, cost of repair, time to repair, etc) Hope this helps in some way Regards, Adam -- Adam Goryachev Website Managers www.websitemanagers.com.au