Oh and with regards blowing away the log files - it would be a really good idea if you change your setup so that doesn't happen again. As I said below, it's nigh on impossible to help debug a system failure after the cluster has been restarted if we have no logs to go on. Cheers, Tim