Pull in 10Gig to the neighborhood over fiber, connect 50 houses with 100m (330ft) of Cat6a UTP for Gigabit speeds.

It's not the fibre that's expensive - it's running new cabling of any type to the premises.

Luckily we don't have either squirrels or beavers here in Australia so our fibre network would be safe (if we had one).

"(But read the fine print, it's 256k up, and absolutely worthless except for 'consuming' their 'premium' offerings.)" Rubbish. Platinum - 100/40 Mbps - http://www.internode.on.net/residential/fibre_to_the_home/nbn_plans/

"Australia will NEVER be a catalyst for technical advancement, the population base is simply too small..." Yep, just like Israel and Sweden couldn't be either.

If your server only takes a few minutes to reboot, then I can see why you wouldn't be so concerned about having to reboot for kernel upgrades. We have Oracle and Sybase database servers that take over 90 minutes to start up all their services (these are 16 and 32 core machines) and not having to reboot them for kernel updates would be a huge win for us.

He's right and he's wrong at the same time.

Oracle calculates its licensing cost based on cores but then discounts that number depending on its calculation of the 'power' of the processor architecture. In the case of x86/x86_64, the discount is 50% meaning that two cores counts as one processor for licensing purposes. This of course means that a license for a quad-core CPU will cost you just as much as two dual-core CPUs.

As for charging you according to the amount of RAM in a server, that's just rubbish. We've just upgraded to 256GB per server in our main RAC and the licensing cost is no different than when we first rolled it out with 32GB in each node.

Try Optimize Google instead. It's a far more actively maintained fork.

Can you explain what version of Blackjack you are playing when you can possibly have 3 x 10 value cards and not be bust?

One here and another here. Both are for older versions (3&4) of RHEL but the same principles apply.

As someone who works with Oracle RAC and RHEL regularly, I'd recommend skipping the shared physical disk completely and using NFS instead. You could (and we do in testing) run the NFS server virtualised as well.

Or even better, Safekeep which uses rdiff-backup but adds easy per-server configurations without writing scripts, LVM snapshots, database dumps and SSH key configuration.