Melting technology and missing black holes
It could only be letters...
Letters It is hot out there. And hot is not good if you depend on bits of metal for a living. This is what Level 3 discovered this week, when its Goswell Road facility took some time out to chill.
This news prompted a flurry of emails. Well, one or two. Telling tales of the heat killing technology all over the place:
Level3's Goswell Road wasn't the only Tier 1 data centre to suffer on Sunday. Globix's rather aged facilities in Prospect House on New Oxford Street were down for a similar period after EDF forgot to put money in the meter. However this shouldn't have prevented the diesel generators from starting, but it appears that Globix may also have had a lapse of memory as they too soon gave up the ghost.
Globix wrote in to 'fess up:
“On Sunday July 23, 2006, Globix experienced a power outage at its data centre in central London as a result of the wide spread power failure experienced by EDF Energy customers across London.
Globix engineers worked hard to restore power to the data centre facility and were able to limit the disruption caused to customers’ through quick action to start the back-up fuel facilities.
Globix is currently conducting a full investigation into the causes of the outage and will inform customers of the outcome.”
And the hot weather wasn't just hitting data centres - a letter from L'Orange confirmed problems with its broadband network:
Orange can confirm that a broadband network outage occurred for a short period in the late hours of Saturday 22nd July. We would like to reassure our customers that only a small number of users on our network were affected. As soon as we were made aware of the problem, we worked with our suppliers to rectify the issue as quickly as possible, and normal service has resumed.
We apologise to those customers affected.
An Orange Broadband Spokesperson
From one environmental challenge to another. This time, the long-waited, seriously overdue WEEE directive looks like it might get an implementation date. At least that is what the government says. A well-placed source is less convinced:
It will definitely happen ... Probably not on the dates you mention ...
There have been three main issues with the delay in UK implementation:
1. No one wants it (including HM Govt): so no one has been in any rush to actually agree it. This still appears to be the case. I will not be at all surprised if July 2007 slips back to the Autumn or even 2008.
2. Historical WEEE (especially that from manufacturers/brands who no longer exist): Manufacturers know that they will be liable for their own mess, but have a lot of trouble with the notion that they will have to pay for mess created by someone else.
3. Not charging Consumers - Initially WEEE was not supposed to cost Joe-Blow any money. Manufacturers were going to do the cleaning up for free! Last night's BBC News seemed to be saying that this requirment had been dropped, and that the WEEE Directive would increase the cost of brown and white goods to consumers.
This will be very well received. Once manufacturers are in a position to estimate the annual cost to them of the scheme, then they can work out how much more money they need to make off a microwave to keep existing margins (yes, it will/could turn into a nice little earner, depending on how well "cost recovery" is monitored ... and monitoring it will not be easy, if it's possible at all).
Other Points: A "takeback scheme" - Consumers taking their own microwaves to "the tip"? This is not how it was originally envisaged (and it will fail if they mean this is the only route).
The working assumption was that retailers who delivered a new washing machine would be required to collect the old one and deliver it into the WEEE scheme. Obviously retailers weren't happy that they incurred a cost, although I assume the "new deal" means they could recover the cost via their delivery charges.
UK Treatment Facilities & Accredited reprocessing/recycling facilities - Blimey! I figured we'd cop out and ship it all to Germany. Be interesting to see who is stumping up the capital to create these. A nice shiny contract for someone (I have an idea who, but I'll keep that quiet for now).
Unless a deal has already been done, this will be harder to create than the statement implies; so far, few entities have been willing to invest given the (low) earnings that were on offer. Maybe things have changed (though I have seen nothing - yet - to tell me that this is the case).
Implemetation Date: In early June 06, the forecasted Consulting Date was Dec 06 and Implementation Date was April 2007. Producers kicked off that this was not enough time. I very much doubt they only had 3-6 months more in mind when they moaned about it!
Tool up. It's wiki versus Britannica. Except that it's not. But don't say we didn't warn you.
In regard to your comment about Britannica being frozen 'until the next version', I just wanted to point out that the Britannica website is under constant revision. Not only do they list the most recently revised articles, but they are about to start an editorial feedback program, whereby registered users will be able to submit feedback on a particular article directly to the editorial team there. While your remark may hold true in regard to the print edition, that would be comparing apples to oranges (i.e. online vs. print), which isn't exactly fair.
I don't want to start a Wikipedia vs. Britannica debate, since that's been done too many times already, but I thought I should bring that fact to your attention, since Britannica is still the most authoritative general reference work on the planet (whereas Wikipedia strives to be much more than that). Beyond that, Britannica is the only project between the two that is appropriate for young learners. I shudder to think of a school-age child searching for any number of four-letter words that are contained within Wikipedia. Or, God forbid, something more esoteric, such as 'felching'. Terrifying...
You took issue with some of the activities accused of making a living from internet fraud:
"There is evidence that [internet] fraud funds terrorism, drugs and people trafficking", eh?
Terrorism I can buy- although it is such a political football I'd rather take that with a pinch of salt- but as far as I am aware, drug smuggling is largely a self-financing operation.
Please think twice before you publish this spin.
So, Google builds its own servers eh? Cormac O'Reilly explained why he thinks his is a particularly good idea, and struck a particularly harmonious chord with many of you. Vive la revolution!
Can I buy one of those servers from Google? Might be an option! Could be a reasonable deal. They might even support the device with some "smart" customer support people. Of course that might be pushing it, then again, they might work so well that the "traditional" support isn't needed. Wow! A reliable computer, that actually is!
Err, scuse me.
How is this *news*? Is it really that hard to think in terms of function and building something that helps you get there? Apparently, google has been doing this for years. Apparently, archive.org even spun off their server building guys into a separate company.
Personally, I don't buy into the big name game that essentially means spending gobs of money on the pretext of handing off all your problems to the hardware guys. Every time the hardware changes a bit or gets shuffled around, and the buzzwords change. You still end up talking to their premium gold telephone that spins an "we're ignoring you" tape.
Shoulda just hired competent people instead of hoping the hardware (or software!) the IT director saw in his in-flight mag would take care of itself.
In fact, it's always seemed to be hype, leaving me frankly amazed that anyone bought into it. Suckers.
The "comment" tag should help you with the first point....
You wanna come over and try to sell that to my boss?
Great story filled with strong supporting examples.
Such a common sense article is long overdue and hopefully reflects a growing opinion of what is capable when "convetional sales gibberish" is forsaken for practical execution in IT strategy.
Thanks for the article.
Jeff IT Consultant
Interesting surname, Jeff...
I have three points upon which I disagree with your article.
Please note that they're all very specific points - the general gist of this article is one that I very, very strongly agree with. Here goes:
1. "Do corporations really need to spend money and productivity on introducing another Battlestar Galactica operating system (Vista) from the Microsoft franchise - one so complicated that it can't quite seem to get out of the shop? Will PC users be measurably better off after the expense and disruption of introducing the next MS Office release?"
Possibly. Microsoft are making some impressive progress with these. Whether it'll be worth the money, whether the software will be stable, well, we'll have to see. At the very least, they're unlikely to have made it *worse*. For reference, I'm an open-source fanboy. And yet I still have just a little faith that MS might've come up with something worthwhile by now.
2. "Why not buy standard networking equipment from reputable commodity providers like Netgear, instead of Cisco - as much as 80 per cent saving is likely to be the only noticeable difference to the business."
Because the expensive Cisco boxes can do all sorts of amusing and exciting tricks with traffic, whereas the El Cheapo kit, um... won't, perhaps? Okay, I must admit that I'm with you on this one, though I'd replace Cisco boxes with an even *more* powerful alternative: commodity computers flying on OpenBSD/pf or one of the many traffic shaping doohickeys that exist on Linux.
3. FOR THE LOVE OF THE IT INDUSTRY, SHUT UP ALREADY! If CIOs and suits get a clue and start demanding this stuff for themselves, us geeks won't be able to implement this stuff in the enterprise ourselves and will lose the oppourtunity to present massive savings and operational gains on our own initiative(s) to the higher-ups. =D
I wholeheartedly agree with you... except about CSN(&Y)! Some of their songs prove that they were also in the cage-rattling business.
Very well put in re: Google's hardware philosophy. I've been saying since the late 90s that this year's server is next year's PC, and that with additional RAM and SATA disks (rather than the rip off that is SCSI) you can build cheaper, more resilient and reliable information services than by relying on the "server" class machines.
Well said that man.
Still with Google, but more to do with traffic issues than who is building what kit. Something got lost in translation here:
I've no idea what "tailbacks" are in the UK, but here in the USA, they're a linebacker in American football (you know, Rugby for wusses).
It boggles the imagination that a driver would need Google to help him avoid running over people playing a sport in (what one hopes is) a closed stadium.
To set your tortured mind at ease, Morely, a tailback is a traffic jam. Like strawberry, but with cars instead.
And now for the really silly stuff. Astronomers can't find enough supermassive black holes. Don't you hate it when that happens? Seriously...
"but astronomers are having trouble locating supermassive black holes in neighbouring galaxies."
Well - the thing about space, y'see, is that it's black. And the thing about black holes...
I bet you'd struggle to find a nerdier letter this week!
Well, you'd think so, and ordinarily, you'd have been in with a shout. But not today:
Well, you see the main thing about black holes is, your main defining feature is, there black.
And the thing with space is, it's main defining feature is, it's black. It's a simple mistake that anyone could make really...
And we'll leave it there, before it gets really ugly. Enjoy the weekend. ®
Sponsored: Benefits from the lessons learned in HPC