Get nagios to monitor each VM, and each host (meh, not so important - only for esxi host logfiles), compared to the ntp server(s), and compare the server against a smattering of external hosts (perhaps including your country's official time standard if you're a government organisation). We're monitoring 250 VMs here, and none of them have ever been more than 0.1seconds out.

I forgot to say: monitor against external ntp providers (asking Networks to punch holes through firewalls for your monitoring host(s) appropriately) even (especially) when using super expensive GPS clocks as your stratum 1 source. You have to remember that GPS receivers are manufactured by cheapest bidder incompetent fools who don't even understand how TAI-UTC works, hence why they're lobbying to abolish UTC time. Symmetricom, I'm looking at you. Good thing leap seconds are updated in the ephemerides at UTC=0, so on the east coast of Australia, are applied erroneously 3 months early when optical telescopes aren't observing the sky.

Don't virtualize anything requiring tight scheduling or a reliable clock, such as a software PBX system performing transcoding or conferencing.

Pffft. We're running cisco's voip stuff one one of their cisco UCS chassis here. Not a problem, and entirely supported.

NTP hasn't been a problem for years, so long as you read and understand the VMware document and have some reasonable knowledge of NTP (more knowledge than the people packaging ntp for redhat, is unfortunately required).

Don't *ever* fallback to local disciplining perhaps except for a single master ntp server for your organisation (if you expect to have an unreliable network to the outside world). Why the fecking hell did Redhat decide that workstations and non-ntp servers would fallback to local disciplining?

Set 'tinker panic 0' in ntp.conf

Get nagios to monitor each VM, and each host (meh, not so important - only for esxi host logfiles), compared to the ntp server(s), and compare the server against a smattering of external hosts (perhaps including your country's official time standard if you're a government organisation). We're monitoring 250 VMs here, and none of them have ever been more than 0.1seconds out.

I've seen some really screwy ntp and time configurations out there (including using cron to rdate to a local server) - it's a pity that competent ntp knowledge isn't a requirement of law. Maybe it's something North Carolina should persue after they've finished passing the sea-level-isn't-really-rising-all-that-much legislation.

That's when I park the new car and go back to driving my '79 4x4.

Is that the one with the "impale driver's heart on crash" steering column option?

Hope so! And that his crumple zone is the boulder lying in the middle of the road instead of a prius.

Google might get a huge kick in the junk over Dalvik, Oracle might get a huge kick in the junk over whether or not...

Could you please stop using the word "junk"? It's stupid.

I could be off-base about the above assumptions, but they are what I've gathered from reading the various articles I could find actually discussing the patents technically and from reading the patent description itself. I'm not a radiophysicist/engineer, so I could be missing something which would be obvious to one.

Knowing the work that these guys do and have been doing for decades now (I was doing radio astrophysics), but not having read the patent, I strongly suspect it relates to the interferometry work they have been doing for the Australia Telescope Compact Array national facility. Interferometry allows (extremely) directional signal detection from omnidirectional antennae, and simularly directional radio frequency interference mitigation.

A crucial part of radio interferometry is doing Fourier transforms. Getting large amounts of bandwidth necessitates doing this in hardware, in parallel. The precursor projects for the Square Kilometre Array mean these parallel calculations needs to be done quickly (realtime), large bandwidth (the frequency range from the sky would ideally be spread over many gigahertz), and massively parallel (terrahertz digital signals prior to data reduction). So need to be done cheaply and in hardware.

The mathematical techniques are highly non-obvious (and extremely neat). These guys pioneered the mass production of the miniaturised supercheap hardware involved. Yes, they outsourced it, but they most certainly did design it all. This all took quite a lot of investment and innovation. The real point of patents.

> They are the kind of company that makes me cringe because I know there are real, legitimate, marketers out there that do use email to engage clients and keep them up-to-date

There are actually clients out there that want to be "engaged"? All the marketing bullcrap I get from Oracle, Sun, Veritas, VMWare etc is because we have a "relationship" with them. Do I want to be engaged? Hell no. Marketers are all just a waste of time. Pity they think they're being useful. All they're doing is clogging up my procmail rules.

Good value! With Internode, on copper ADSL2+ (24 down, 1.5 up), 150GB monthly quota, all for... $50!

I'll let the early adopters adopt this one. (on the other hand, those poor sods that hadn't heard that you didn't need to use Telstra would probably consider this a good deal).

I hope the Plank Constant is not found to vary over the life of the universe, as alpha has been conjectured to.

I am a little surprised that the several spheres of silicon scattered around the world hadn't already redefined kg standard. I saw one of those balls 10 years ago, and understood then that the work was almost complete - the deviations from a perfect sphere were negligible, radius well determined, and purity excellent.

I'm also a little surprised that these versions of the kg standard need exist at all. I thought it was exactly 1L of water. Which is exactly 10cm*10cm*10cm of water. And a metre is exactly the distance light travels in 1299,792,458 of a second (it had formerly been 1,650,763.73 wavelengths in vacuum of the radiation corresponding to the transition between the 2p10 and 5d5 quantum levels of the krypton-86 atom).

I guess "water" isn't sufficiently well defined or reproducible.

'Kandlbauer "accepted his fate in a manner that commanded great respect from all of us'. What? It's not fate when it was his own stupid choice to climb a high voltage pylon.

Why was he allowed to drive despite his arms being controlled by an unreliable experimental medical/machine technique? His motto was "Don't live for others, live for yourself!". Meanwhile, he was always far more likely to contribute to other peoples deaths because he was allowed to drive.

I just hope he didn't hurt or kill anyone else when he crashed.

No, the youths are turned off crappy classical muzak played through shitty speakers. I hate it to. I always know when I'm in the far eastern suburbs of Melbourne when they play some horribly piped noise through horn speakers designed to mangle train announcements and not for high fidelity.

When shops play similar, I make it my aim to get the hell away from such places. I assume they didn't want to sell me anything.

The great thing about a serial console is that it doesn't take long to figure it out. And you only need 3 wires to get there.
Another nice thing about it is that it's point-to-point, so you don't have to worry about your signals getting lost.

Heck, you can create a serial interface from discrete components if you're really into fun.

Wow. Miss the point completely. In the datacentre, trying to configure my SAN, I don't give a flying rats about whether my cable only needs 3 conductors and I can build a device using only discrete components. I'm not building devices and the cable came in the box with the device. I want to plug my special magical cable somewhere into a special magical and standard port somewhere on my laptop (ie, not a serial port), and have it talk to a special magical port somewhere on my device. I'd rather it be error detected and corrected just so that when partitioning my device, it didn't interpret "create new partition" as "wipe all partitions".

I strongly suspect a pl2302 or similar usb-serial chip that has linux drivers only costs a few cents, and the USB communications are error corrected (and the signal lines from the converter chip to the internals are all done within the metal enclosure of the device I'm configuring, so should be fairly resistant to errors). So if these devices were built included as standard instead, I'd have a much better chance of getting my data onto the device error free for some time in to the future until USB has been superseded. I've got devices at work that were advertised as containing "USB interface", which instead came with USB serial converters. They work fine. Just add udev rules to match the device and create a symlink somewhere in /dev, then configure minicom to talk to that location.

Of course, I'd be equally happy with ethernet (unencrypted telnet talking on some random private IP would be fine, this port need not be plugged into the network) - I'd configure my laptop to send all private subnet ranges to the ethernet port that was plugged directly into my device (if the laptop needed network access, have a second port or wireless).

If you get fooled by politicians who lie by abusing statistics, then that's a pretty good sign you don't undestand statistics and need to learn more about it.