(Credit: Memory stick 3
by Ramasamy Chidambaram, Royalty free)

The data loss scandal du jour is the disappearance of a hard drive containing details of 1.7 million people who had enquired about joining the UK military.

The story ticks all the boxes: the data was unencrypted, it included sensitive information such as passport numbers, and it only became clear it went missing some time after the fact.

As ever, it seems that human error rather than technology failure is to blame. This isn't unusual; people have been misplacing documents in offices since the dawn of enterprise time, and it's only the sheer scale of information that can be crammed into a hard drive or USB key that makes the whole process so utterly scary.

A recent survey by RSA casts an interesting light on the problem. The company asked 417 attendees at three recent conferences a series of questions about security policy. When asked "Have you ever lost a laptop, smartphone and/or USB flash drive with corporate information on it?", 1 per cent said they had done so frequently, and 9 per cent said it had happened to them "sometimes".

That indicates there are at least four complete klutzes out there who are continually misplacing portable storage, and who we can only hope don't work in data security. But the more disturbing aspect is the fact that, in effect, 10 per cent of information is likely to be misplaced.

Companies routinely take out insurance to deal with relatively unlikely risks. I suspect if I told my insurer that my house had a 10 per cent chance of being flooded, they'd either laugh me out the door or demand that I take some precautions to minimise the damage.

The same logic should apply to portable devices and the information stored on them. Can you hear the pigs flying overhead?

Comments (2) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• UK prisoner data goes missing
• ATO admits staff have lost data, sent porn e-mails
• US military to use network warfare to break enemy
• McAfee buys data leakage firm

Google's Sergey Brin
(Credit: Google)

In a post on its official blog this week, Google has once again talked up the environmental efficiency of its datacentre initiatives.

According to Google, its data centres are so efficient that your own PC uses more power while waiting for a query result than it does in processing that query. "Google-designed datacentres use nearly five times less energy than conventional facilities to feed and cool the computers inside," Google operations VP Urs Holzle wrote. "Our engineers worked hard to optimise every element in the datacentre, from the chip to the cooling tower."

I don't want to suggest these aren't worthy things to be doing. But there's one piece of information that's still missing from this equation: just how many datacentres is Google running? Is this technology being deployed in a dozen locations? A hundred? A thousand?

Google isn't saying, in line with its long-standing policy that datacentre numbers are the kind of secret detail that don't need to be shared with shareholders, governments, energy providers or the general public.

I've whinged about it before, but the fact remains: this message would be a lot more effective if there were some hard numbers behind it, and it creates a dangerous precedent where corporations expect us to trust that they're doing the right thing for the environment without actually detailing the impact their operations have.

Like Google's similar plan to launch oceanic datacentres, the devil is in the detail.

Comments (0) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• NSW govt consolidates shared services
• Oracle unveils first hardware product
• Datacentre 2020: Data security gets physical
• Micron unveils server-grade SSDs

SanDisk next month will start offering a 16GB microSD card, which — unsurprisingly given the format — is aimed at the mobile phone market. The SD card will set you back $141, and will initially only be available in Crazy John's. (There's also an M2 model for Sony Ericsson models, but I'm not sure how many people will need to get that niche.)

That amount of capacity inevitably leads to the question: how are you supposed to fill up that space? Pictures might be one potential answer, but 16GB will give you a hell of a lot of semi-quality phone shots. Indeed, the whole Pictures folder on my PC doesn't take up that much space.

The other obvious candidate would be video, but again, it's debatable that you need that much space for video that's viewable on a phone screen. My 8GB iPod Touch can handle an entire US sitcom season and still have plenty of room to spare.

Not everyone will agree, of course. For its announcement, SanDisk even rolled out a tame analyst, Avi Greengart from Current Analysis, who made the following observation: "Flash memory cards have increased in storage capacity, but even an 8GB card may be too small for anyone with GPS map data, a few movies, a game or two, a presentation file and other applications."

But if this is you, then I suggest you re-compress your movies into a more screen-friendly alternative and stop overloading your PowerPoint presentations with meaningless graphics. Save the capacity for where you really can use it — on a desktop PC where the OS will reclaim it in the blink of an eye.

Comments (21) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• Apple iPhone 3G (16GB)
• First Take: SanDisk Wi-Fi SD Combination Card

"I often refer to encryption as crypto fairy dust," Eric Hibbard, chair of the Security Technical Working Group in the Storage Network Industry Association, said in a recent interview. "A lot of IT managers sprinkle this on and think it makes certain problems go away."

The reality, Hibbard suggested, is rather different. "If you're doing encryption in the storage ecosystem, the pay off is very limited. A hard drive or tape drive wandering off is a real problem, but that's not a data confidentiality issue; it's a media confidentiality issue. If you're talking about sensitive information, encryption is just one tool in the toolbox. If you don't have that mated to tight authentication and access control, you're screwed."

Of course, there are plenty of reasons why such a mating isn't happening. Getting to that kind of integrated nirvana is a worthy goal, but rarely happens in IT environments where heterogeneity is a fact of life. There simply isn't time, budget or staffing expertise to bring it all together, so access control tends to be limited to the most pressing projects.

Do you think Hibbard is on the right track here and it's time to vacuum up the fairy dust, or is encryption still the best option of a messy bunch for basic data security?

Comments (7) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• Mac users to get PGP whole-disk encryption
• Skype encryption too tough for German police
• Cisco and RSA to offer SAN encryption
• Defence encrypts whole laptop fleet

(Sandwich,toast by Oscar Martinez, Royalty Free )

Brennan IT certainly didn't anticipate selling culinary features when it began building its new datacentre in Brisbane earlier this year. The centre was designed to host both Brennan's own telephony management equipment and co-located servers for commercial clients, so there was a kitchen, toilet and shower facilities connected to the staff side of the building.

However, the kitchen turned out to be an unexpected selling point for external clients as well, as account manager Sandy Forster explained when I toured the centre recently.

There's a fit-out room on the premises where visiting techs can make sure their equipment is working properly, and several clients suggested that offering access to the kitchen would be a useful bonus. On the original floor plan, the two were separate, but after the idea was raised, adding a doorway meant that budding data centre gurus could also show off their inner Jamie Oliver.

Aside from the minor hassle of needing a security policy to stop visitors nicking staff member sandwiches, this seems like a simple and obvious step, but such features are often low on the list. Reputedly, some large datacentres don't even have toilets, since they're supposedly meant to be "lights out" facilities. Bet that makes for an unpleasant carpark.

In the case of on-premises data centres, many are located in odd nooks of the building and nearly all have high levels of security by corporate standards, so access to amenities is often awkward. But if you want to show staff that they're as valuable as the data they look after, offering a fridge and a microwave is a good way to start.

Comments (4) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• CenITex needs to be governed for success
• VMware reveals ESX 4.0 features
• Google's ocean datacentre plans a tad fishy?
• Storage and server superguide — all you need to know

While it was released earlier this year, XP SP3 has only recently started being pushed out to Australian users as part of Microsoft's Windows Update service. If Windows Update can install the update successfully, then you're typically looking at 70MB or so of downloading — painful, but not impossible. If the install chokes, though, Microsoft recommends downloading the full version of SP3, which is a rather less cosy 316MB.

For a corporate, that's not necessarily a problem, but for individual consumers on a slow connection, or with a minuscule download limit, it's a major pain. In the past, Microsoft has offered to send out CDs containing large patches, and it is doing that for SP3. What's shocking is the cost: $14.95 for a CD that will apparently take between two to four weeks to arrive at an Australian address. At that speed, even a dial-up download might be quicker.

When I asked Microsoft how it could justify this ridiculous state of affairs, I got the following reply: "Microsoft manages duplication of its software, including Service Packs, at regional duplication facilities. As such, the copy of Windows XP SP3 you are ordering is coming from Singapore, which reflects the shipping costs you've referenced as well as the time to delivery."

That partially explains the shipping delay, but it's a lousy excuse for the price.

Australia Post charges $7.75 to send a CD-sized package to Singapore, so why does sending one the other way cost twice as much? It's not likely to be materials or handling cost; the last time I checked, blank CDs cost rather less than $7 each, and if Microsoft can't manage the data processing from an online form in a cost-effective way, it probably should get out of the software business altogether.

Service packs shouldn't be an excuse for profiteering, but it's hard to view this particular exercise as anything else — and it's a lousy incentive for people to keep their machines secure.

Comments (29) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• Microsoft: XP SP3 won't arrive until 2007
• XP SP3 coming this month?
• XP SP3 puts AMD users into endless reboot loop
• Windows XP SP3 upgrade spells trouble for IE

A sketch from Google's floating datacentre patent.

On the one hand, I can see how having a datacentre in a location where only seals can visit it would help to keep it a secret. On the other hand, it's going to be a bugger getting people there for the initial fit-out — and what if the centre gets held hostage by high-tech pirates?

We're always being told about how hackers have turned into professional criminals, and there's no reason to assume that a little bit of salt water would put them off.

What makes this story really weird, though, is the assumption that ocean-based power might suffice in an era when getting sufficient power to datacentres in areas with massively well-developed infrastructure is proving increasingly problematic.

Sure, there won't be a lot of krill queuing up to charge their ultra-compact iPod Nanos, but it hardly seems like a practical scheme. As one datacentre manager said when I mentioned this to him today: "Some people just have too much time on their hands". Or maybe Google is drawing inspiration from the infamous UK pirate radio stations of the 1960s, which escaped British regulators by running offshore.

Comments (4) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

While I'm wary of placing too much emphasis on statistics that only involve 50 locals, two figures did stick out. Firstly, the number of C-level executives on corporate DR committees has dropped from 50 per cent to 33 per cent.

Perhaps we need some nastier natural disasters or terrorist attacks to refocus everyone's attention.

Secondly, the percentage of applications defined by the surveyed businesses as "mission-critical" is now an astonishing 64 per cent.

I'd venture to suggest that if you think two-thirds of your IT is mission-critical, you're either running an incredibly lean and efficient operation (not likely given the universal complaints about keep-the-lights-on budgets) or you haven't got a clue how many applications you have and which ones you need to manage (all too likely).

A disturbing sub-trend pointed out by Symantec's Australia and New Zealand director of systems engineering Paul Lancaster is that many businesses would define a core application as mission-critical, but not accord the same status to underlying systems (like databases) needed to keep those systems running.

Given that, it's not surprising that when these businesses ran DR test scenarios, a good third of them failed.

Cluelessness is also evident in the rush to use virtualised systems as a cheap-and-cheerful DR option. "Virtualisation has sprawled instead of grown," Lancaster said. "People have seen apps being able to be put on a single server and the management side of it has been more or less left up to the application owners."

That's a real (non-virtual) disaster waiting to happen.

Comments (1) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• St George revamps data warehouse
• CommBank chops IT spend
• Virtualisation saves $1.8m for Corporate Express
• White House loses millions of war e-mails

I've just been on a brief trip to India, where tendering for absolutely everything seems to be a way of life, at least if the advertisements in the newspapers are any guide. Whether it's a business title or a mass-market tabloid, every other square centimetre is filled with invitations to tender for everything from freeway construction to sewerage removal. Combined with endless memorial advertisements for long-departed relatives, it creates a print ambience, which is nothing whatsoever like the Daily Telegraph.

As a professional storage watcher, I was naturally curious to see if IT infrastructure was part of this trend. For the most part, the answer seems to be "no". One potential explanation is that such advertisements only appear in specialist publications, but that doesn't account for the diverse range of options that are otherwise being sought. Another possibility is that with so much basic infrastructure to build for an expanding population, IT for public works is somewhat lower down the scale.

One tender for a "disaster management system" for Indian Railways initially held some promise, but turned out to be for computer training systems to help staff deal with higher volumes of traffic and the accidents that seem to inevitably follow — not an area I think EMC or IBM will be clamouring to fill in a big hurry. (Though I was taken by this interesting disclaimer: "This is a working day but if subsequently declared a holiday for any reason, the tenders will be opened on next working day.")

Indian Railways did, however, have another tender relating to storage and document management, albeit on an astonishingly basic level. Tender 197-S-104 seeks bids for "Provision of Portable Cabin for record maintenance at C&W wing of DRM office".

Now, this record maintenance does seem to have an electronic component, since the provision of a desk with keyboard access and room for a monitor is a core requirement of the tender. However, there's more information on the kind of paint needed for outside the building and the specifications for venetian blinds than there is on the technical side.

Attention to detail is important for any IT project, but when it comes to specifying office layout, most IT managers are lucky to get consideration for anything other than security and cooling issues. You probably don't want to have this kind of consideration added to your daily task list, but it's something to bear in mind if an Indian IT services company starts head-hunting you.

Comments (0) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• Conroy defends 'fluffy, insubstantial' FTTN tender
• Auditor: Labor in the clear over FTTN tender process

A month after the first iPhones hit the streets, I still find it jaw-droppingly amazing that people were willing to sign up more or less without any consideration of what plans they were going to have to use.

Early adopters have paid the price, as my colleague David Braue points out, since the plans have since become somewhat more reasonable for people who were willing to hang around and wait for the market to settle in. Apparently the iPhone-obsessed aren't suffering from the economic downturn.

I've no intention of buying an iPhone myself (I like having a real keypad), but between writing about it and watching endless debates about purchasing decisions on Twitter, I've encountered a fairly heavy volume of discussion of the buying process.

And while debate over the appropriate plans might have been embarrassingly minimal (witness all those people who flung $100 towards Optus without even knowing what was being offered), there have been two more basic points of debate.

One is which colour iPhone to purchase, with black generally being favoured over white (whether for cool factor or a sensible recognition that white gadgets end up looking mucky, I'm not sure). The other aspect — more germane to Snorage's view of the world — is whether it's worth coughing up the extra money for a 16GB model, rather than the entry-level 8GB option.

Early sales reports suggest that the 16GB model has proved more popular, with some buyers begrudgingly accepting the 8GB model in order to get their hands on a phone sooner.

I'd suggest that this has very little to do with the iPhone's performance as either a phone or an applications platform, and everything to do with its legacy role as a music player.

Even 8GB is enough to store more music than anyone's ever likely to listen to, and it's at the high end of what's on offer in the next-down Nano range. But it's still only a small fraction of what's available in a disk-based iPod classic.

If you've managed to fill up an 80GB model, then even the 16GB model will require you to make some musical sacrifices. (Add video into the equation and you can fill everything up much faster, of course).

The rising capacities of flash memory would seem to make it a safe assumption that future iPhone models may offer even more storage. On the other hand, if the data plans keep getting cheaper, remote access via the cloud might prove more appealing than whacking in some more memory.

In the corporate world, of course, more storage is not necessarily a good thing in this context. Despite the addition of Exchange support, the iPhone is still a lousy choice of smartphone for business users, since it's harder to manage and standardise en masse.

That doesn't mean that pushy executives aren't going to demand support for it, of course. The only comfort IT managers might be able to draw is that an 8GB model can only store half the volume of sensitive corporate information, though that's plenty enough information to cause a major embarrassment if it falls into the wrong hands.

Comments (7) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• Much ado about iPhone kill switch
• iPhone nano reports surface
• Virgin iPhones already sold out
• Virgin Australia to launch iPhone?

In a posting on the Microsoft storage team blog, engineer Kevin Allen discusses an apparent bug in the Robocopy folder replication utility included in the Windows Resource Kit for Windows Server 2003 and XP and then added as a standard feature to Vista and Windows Server 2008. (Sadly for techno-fetishists, the robo part stands for "robust" rather than "robot".)

(Credit: GiniMiniGi)

When replicating information, the original Robocopy will mirror any files that have changed their content, but won't replicate files which have had their security permissions changed but are otherwise unaltered. That might look like a flaw, but according to Allen, who wrote the basic software more than a decade ago, "this behaviour is by design". While that kind of phrasing is sometimes an after-the-fact justification for a product design which deserves punishment, in this case Allen makes a pretty compelling argument.

Firstly, not checking every file for pure security alterations makes the replication work more quickly. Secondly, Allen argues that most companies are happy with directory level security, and that setting file by file permissions is the kind of fiddly nitpicking most people won't bother with.

Finally, if you do want to replicate on this basis, you can use a chain of commands to ensure that all files altered with security do get changed as part of the replication process. The option is there if you really want it, but as a default behaviour, it wasn't necessary.

Of course, one thing Allen probably couldn't have anticipated was that Bill Gates' "security before all else" dictum for Vista would eventually mean that aspect of Robocopy's behaviour would have to be changed, no matter how much it slowed things down. (The fact that Microsoft is still promoting User Account Control as a good idea shows how doggedly it has clung to that view, but I digress.)

Hence in the Vista and 2008 versions of Robocopy, there's an option to ensure security changes are replicated by using the SECFIX switch on the command line. This isn't a bad compromise: the default command still has reasonable performance, but there's an option for people who regularly find themselves messing with security settings for whatever reason.

The downside of this approach is that it means Robocopy doesn't behave consistently between different platforms, and scripts written to run on Vista systems won't make any sense on older machines.

In the Microsoft purist world view, that wouldn't matter, because you'd have Vista and 2008 running everywhere. In real-world heterogeneous environments, it's just another potential source of pain. Maybe storage administrators do need a slightly masochistic streak after all.

Comments (5) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• Scott Charney: Microsoft's security chief reveals all
• Queensland plans huge Exchange 2007 project
• Vista is "New Coke"
• Windows chief jumps ship for Juniper

While users of Web-based email services such as Gmail often like to boast about the massive amounts of data they're able to store, that doesn't actually eliminate the problem of how to manage the storage systems needed to support that — it simply shifts them into the "someone else's problem" category.

At Google's recent developer day in Sydney, Daniel Reyes, the head of engineering at MySpace Australia, outlined how MySpace was dealing with its own message storage problem.

While you probably know MySpace primarily for its endless store of "friends" for teenage relatives and/or as a neat means of listening to music online, it also supports what turns out to be a pretty considerable messaging infrastructure.

According to Reyes, 160 million messages are sent and 300 million received by MySpace users daily. At peak periods, that translates into 20,000 messages a second. That information is stored in 400 mail databases, each of which is around a terabyte in size.

Naturally, searching and accessing that volume of data can be a challenge. One of the solutions MySpace has adopted is Google's Gears technology, which allows localised storage of information accessed via an online environment — that is, getting MySpace users to store some messages on their own hard drives rather than in the cloud, with Gears ensuring that the two systems are blended together effectively while visiting the site.

This isn't a one-size-fits-all approach. While MySpace has around 380 million users, only 110 million are considered active, and of those, only a small percentage — those with a high level of messaging activities — have been prompted to install Gears and localise their mail store. MySpace originally targeted users with more than 5,000 messages stored, and more recently expanded its approach to cover anyone with more than 2,000.

Reyes said that the company is also considering localising the search system for friends on the site, which could prove a more challenging project given the MySpace culture of "more friends = better".

Getting users to take a hand in the management of data is often a useful step, although, as I've noted in other contexts, there's frequently resistance to this kind of change.

And sometimes there are people who are almost impossible to satisfy. "We had one user with over 100,000 messages," Reyes said. "One of the issues he had was that it was taking too long to replicate." Clearly the time taken to accumulate that amount of social detritus wasn't an issue.

Comments (4) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• Will you manage in the exabyte era?
• Virtual servers? Untangle your cables first
• Many mail make managers manic
• Wanna save storage? Try social networking

I was reminded of this during the week while hanging out at the Hitachi Data System SAN Technology Centre in Odawara and talking to Hubert Yoshida, HDS' chief technology officer. Yoshida predicts that storage volumes are about to cross an important threshold.

"We expect that within the next three to five years we will have a customer with an exabyte of data," Yoshida said. "Today we have customers with close to 100 petabytes."

What can you realistically do to manage a million terabytes or more of information (or 100 petabytes for that matter)? Whatever the answer, it's unlikely to be found in current systems.

"Even though in storage we have been doubling in capacity almost ever year, the basic architecture of storage systems is in many cases 20 years old." That approach won't be sustainable in the exabyte era, Yoshida suggests: "It requires a new approach and architecture. We have to have a fundamental change in the way we do architectures and implement technologies. It's not just a matter of getting bigger disks, we have to change the way they work together."

A critical area for improvement is interconnect speeds. "As we move towards exabytes of storage, we can't afford to move every byte across the network," Yoshida said. As well as faster pipes, intelligent processing (such as de-duplication and compression) can help simplify that problem.

In the short term, one useful strategy is to focus less on three-year storage capacity plans, Yoshida suggested. "Instead of buying all this big storage today, buy what you need as you need it — because capacity always gets cheaper."

Angus Kidman travelled to Japan as a guest of HDS.

Comments (0) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• CSIRO in three petabyte storage upgrade
• Taking datacentres on the road
• Nanotech points way to petabyte disk drives

Yet shoving everything into a hosted environment does effectively create a quick and dirty disaster recovery strategy. If most of the services you need can be accessed via a conventional PC with an Internet connection, then the inability to get to your main premises might not require failing over to a backup location.

I was reminded on this during a recent lunchtime meeting with Michael Bishop, the general manager of Queensland managed speech services company SmartSpeak Solutions.

Disaster recovery certainly wasn't on Bishop's mind when SmartSpeak, which he founded in 2005 and now provides speech services to 1,300 customers, decided earlier this year to implement an SaaS-based customer relationship management (CRM) solution to help manage calls to the company's Network Speech Operation Centre (NSOC).

The basis for choosing SaaS rather than a locally hosted solution was simple pragmatism: Bishop foresaw an inevitable interruption to the business of months if a conventional software package had to be installed. Even with the RightNow solution which the company eventually chose three months ago, Bishop advocated a limited roll-out, making the software available initially to just 12 employees.

SmartSpeak now intends to expand that to other divisions, though the process is still going to be gradual; Bishop noted that many reporting features hadn't even been switched on until recently.

Since SmartSpeak is in the business of selling managed services itself, Bishop was familiar with the main arguments used to promote the software services approach, which included removing the need for customers to do their own backups. But the disaster recovery benefits of its CRM implementation were an added and unexpected bonus.

"There's been a huge advantage to us from a disaster recovery perspective, but that was something we completely overlooked during the planning process," Bishop said. "That for us was a big benefit."

Off-premises access hasn't replaced the core DR policy which SmartSpeak developed and includes access to a failover facility if necessary, but it does provide a useful second layer. It also makes it easier for staff to telecommute as needed; Bishop noted that some employees now preferred to work from home early in the morning and avoid peak hour traffic jams.

Of course, all this is very much reliant on the service provider offering a solid quality of service. "With any managed service, you don't know what is in the cloud, and there's a certain level of trust," Bishop said. But if you're trying to sell management on the benefits of the SaaS approach, DR is certainly a useful extra weapon to have in the armoury.

Comments (1) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• Enterprise Storage FAQ
• WebCentral arms itself for storage explosion
• Quantum leap in storage for Deakin
• Storage: The inside story

One of the less obvious disadvantages is that apparently the famous are far too busy (probably being manicured) to remember to back-up or to organise their data.

This is a topic which Snorage has touched on a couple of times before. Back in April, we looked at the sorry case of Scottish pop star Calvin Harris, who left the only copy of his latest album on a laptop in his checked baggage, which subsequently encountered the luggage-swallowing vortex that was (and perhaps still is) Heathrow Terminal 5. There's been no word yet on whether Harris ever got the bag back, but hopefully he's at least downloaded a copy of SyncBackSE.

Then there was Hong Kong star Edison Chen, who made a more common storage mistake: keeping the data in the wrong place. Sending in his Mac for repairs when the machine was jam-packed with saucy shots of Chen and various starlets proved to be quite literally a career-killing move.

Really, you've got to hope the sex was good, given that the whole scenario could have been avoided if Chen had kept his private pics on an external drive, a wise strategy for anyone who has ever gotten naked in front of a camera.

And it's not like he was the first celebrity to get his genitalia unexpectedly exposed in this way. Back in 2005, a sex video featuring Limp Bizkit singer Fred Durst began doing the rounds online.

Durst sued sites that published the video, arguing that it had been stolen from his computer by a hacker. While the mere use of the word "hacker" suggests a greater degree of technical sophistication than the average rocker, encryption on a separate drive probably could have avoided the whole situation.

Once you start looking, celebrity doom scenarios turn out to be rather common. A sad creative fate befell Cornelia Frances, who has played Morag on Home & Away since more or less forever, and who also had a stint hosting the Australian version of The Weakest Link. After penning her autobiography, Frances planned to emulate former cast mate Judy Nunn by writing a novel, but fell victim to robbery. "They took everything, including my laptop," she told Digital Spy earlier this year.

"I was actually writing, I was doing two different books that were on that laptop that I can never replace. So that's gone. I'll have to start again but there you go, we have to live with that." Actually, Cornelia, you don't; you have to get a backup drive pronto.

Lack of money clearly isn't an issue (especially at the price USB sticks go for these days), so we might have to blame the star-studded lifestyle. Given the choice between some simple file management and attending the opening of an envelope, the latter is always going to win, it seems.

But what are the big picture lessons here? Nothing that a storage manager hasn't repeated until their lungs bleed: develop a backup strategy and stick to it. Keep important data backed up in multiple locations. And store data in a location consistent with its sensitivity.

Stars already have stand-ins so they don't need to lurk in front of cameras for hours on end. If they can replicate themselves, surely they can replicate their hard drives.

Comments (1) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• Laughing your way through data disasters
• Top 10 worst IT disasters of all time
• D'oh and un-d'oh: 4 disaster recovery solutions
• Disaster recovery essentials

I was reminded of this possibility during a presentation by Chris Wickham, customer services manager for Wellington City Council in New Zealand, at the recent ITSM conference on the Gold Coast.

While local councils in the Antipodes don't (to be brutally frank) enjoy a reputation as cutting-edge deployers of advanced technology, Wellington has been working hard to update its systems, and a shift towards more portable technology has been a key element in that strategy.

"There's a general trend of people having PDAs and memory sticks as replacements for laptops. Laptops are so 90s, aren't they?" he noted. The ability to store crucial business documents, and possibly even a log-in to replicate your main work environment, on a USB stick helps eliminate the "my fingers are too much like sausages to use this phone" argument that often gets wheeled out by sweaty executive types.

Wickham recognises the potential, but says there's been a considerable downside to the council's move to using a mixture of phones and USB sticks.

"It sounds so tempting: dump your notebook and just travel with a smartphone, possibly supplemented with a USB key ready to launch you onto your corporate desktop," Wickham said. "But what kind of nightmares does that create for the team in the back room? It's been difficult."

Some of those challenges are perhaps inevitable given a radical change in infrastructure. Others are exacerbated by the always-on nature that a phone-and-USB combo provides.

"We've had staff on holiday in London being rung up in the middle of the night about some minutiae at work," he said.

None of this is necessarily an argument against making the switch — after all, laptops have their own support issues too. But the assumption that the simplicity of operation that plugging in a USB stick provides can be replicated throughout IT is a dangerous one. Proceed with due caution to avoid a sticky ending.

Comments (1) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• Record breaking warehouses break usefulness rules
• OS religion almost dead in the datacentre
• Many mail make managers manic
• How do you deal with 250,000 tapes?

I was reminded of this during a recent interview with Martin Fletcher, the chief operating officer for the Australian operations of Mortgage Guaranty Insurance Corporation (MGIC), which set up shop down under in 2006 and provides insurance for loan companies against possible credit losses.

Right from the beginning, Fletcher was determined to set up a data warehouse to allow proper tracking of the business' performance.

"I've set up companies previously and it's very clear that to get to the stage where you can analyse the data, you need it all in one place," he said.

MGIC didn't want to simply import the technology used by the American parent company. "We're a subsidiary of a very successful US business, but as we looked at systems for our home office team, they've got something that suits a business of 3,000 people and we've got 200. The scale of it was overkill for what we needed."

Effectively starting a new company meant that no existing systems had to be dismantled. "From day one, we've hired people in the team here who are very keen to design their own processes from scratch," Fletcher said.

However, that didn't mean that the familiar tangle of spreadsheets and individual documents was entirely absent. Because the company had to be in operation from the word go, individual monthly reports had to be generated until the data warehouse was up and running.

That scenario was probably unavoidable, according to Fletcher. "Realistically, we couldn't have started it any earlier. Everyone's delivered the pieces of the project on time."

Fletcher originally thought that defining the long-term requirements could be carried out in-house, but MGIC IT leader Joe Gullotta convinced him that some external advice would be useful.

"My initial thought when I talked to Joe was we would be able to define our requirements, but early on Joe was advocating using consulting experts. When you look at your business, you may not understand the best way to organise your data to analyse it." MGIC eventually engaged Altis Consulting to help plan the system.

Accessibility was a key requirement. "As we started thinking about the future capabilities of the BI tool, I wanted it to be very user interactive. I didn't want a team of reporting analysts who were the only ones who could use the system."

Even with the basic architecture in place, planning for future needs was tricky. "The biggest challenge was we didn't have a lot of raw information in our portfolio because at first, we didn't have a lot of loans insured. Often we were describing qualities about the portfolio which we didn't have any data on. That was one of the problems we had to overcome."

With the MGIC project now well-established and able to deliver daily analysis on a wide range of metrics, what advice would Fletcher offer to other companies facing similar challenges?

"Definitely spend time upfront thinking about what you're trying to achieve, not just today's requirements. If you work closely with the vendor you can get a vision for what's possible in the future. Distinguish bells and whistles from essential things to drive the business forward. We spent a lot of time thinking about what we wanted as minimum. Then once we met with the vendors we had a much more realistic discussion."

Comments (0) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• Virtually large but apparently small
• Record breaking warehouses break usefulness rules
• OS religion almost dead in the datacentre
• Many mail make managers manic

I was reminded of this at the recent Gartner emerging trends roadshow, where analyst David Cearley discussed disruptive technologies and how to work around them.

Part of the reason we don't get into long-term planning, Cearley argued, is because continuous improvement means we simply haven't had to bother.

"Moore's Law is the crack cocaine of IT," he said. "We can deal with increasing data loads because stuff just gets faster."

That scenario is ending, Cearley suggested, because the shift to multi-core processors hasn't yet been reflected in software specially written to take advantage of them: "There are some written with a certain amount of multi-core capability, but not many in enterprise applications."

Storage generally falls into this area — while we may be very concerned with data throughput at a hardware level, when was the last time you benchmarked the basic software operations of your system? And does it contain any multi-core code? I suspect the answer to these would be "never" and "no", phrases which seem oddly resonant with the generally down-market tone of this week's column.

So let's get back to the complications. "The result is you can go from one core to 16 cores and those applications don't get that automatic power boost. You may even see a decline."

The bottom line? Powerful systems no-one can properly use, and a lingering sense of dissatisfaction that we're not getting everything we could be.

The solution is rather less sexy, but ultimately less painful: planning.

"Start doing some inventory to identify which areas will have potential problems," Cearley advised. "Today, you're not going to be feeling a lot of pain. In four years, there'll be a lot of pain. It's not that in 2008 you need to spend a lot of time rewriting everything, but you need to spend some time planning for the long term."

Comments (0) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• What would Dr Who do?
• Sticking it to USB sticks, again
• How do you deal with 250,000 tapes?
• Many mail make managers manic

In an era where disk costs drop faster than the soaring costs of petrol, it's often easy to forget that there are large volumes of data still sitting on older media. Petrol is a good case in point for why we still need this stuff.

The rising cost of oil potentially makes some of this data more useful than ever for mining companies: oil reserves which once looked uneconomic might now be worth exploiting, and old survey data may well be worth revisiting. But what do you do when that's stored on an ancient backup tape, no-one has the relevant drive anymore and you're not even sure what format the data is in?

Spectrum Data specialises in both long-term storage of archival media and in helping companies pull data off those tapes when the need arises (be that a new need for analysis, an unexpected lawsuit or just a realisation that a bunch of older tapes could be stored more compactly on new media).

Some of this stuff is historic, not just useful. Founder and director Guy Holmes shows me an original NASA tape with data from the Apollo 11 mission, which is suffering from an interesting problem: glue from the label leaking into the centre of the tape reel. (Don't panic; it's a fixable problem and it's not the only copy.)

While we're used to the notion that tape is easily damaged by water, temperature variations and poor handling, other factors come into play, as Holmes explained during my visit.

One that we don't think of much these days is the brand of tape.

While consolidation means there's now only a handful of manufacturers producing magnetic media, and all reach a pretty good level of quality, that wasn't the case in the 1970s and 1980s, Holmes said.

Some manufacturers would sell off inferior batches of tape which would then be rebranded and sold at bargain prices. At the receiving area for tapes at Spectrum, there's a poster detailing dozens of brands and rating them from "excellent" to "poor". Excellent tapes can be read pretty speedily (once they're matched to the right machine); poor tapes may need hand checking and baking (to remove excess moisture) before being read very slowly and under constant supervision.

Even assuming you can read the data, there's no telling what format it might be in. Labels on archive tapes are often deceptive; some businesses would print up labels in advance and use them even if the contents turned out to be different. It's enough to make your reels spin.

Comments (2) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• Crikey, Calvin, what were you thinking?
• What would Dr Who do?
• Sticking it to USB sticks, again
• Burning down the warehouse

I was reminded of this during the week while chatting with Marty Gauvin, the managing director of Hostworks, which specialises in enterprise application hosting. According to Gauvin, Windows and Linux boxes have a roughly equal share within the company's datacentres, and decisions made on platforms now tend to be driven more by business need than OS obsessions.

"The religion is going out of it. You really can't pick which way a customer would go," he said in an interview at the company's Adelaide headquarters earlier this week. "It's much more a platform-based decision."

That wasn't always the case, Gauvin notes.

"Initially, people's expectations of the two systems were quite different. It used to be you'd never run beta software on Microsoft boxes; you'd always run beta software on Linux. We had one client who every morning would want the daily build of Tomcat installed. But what's actually happened over time is the two platforms have got a whole lot closer."

That doesn't mean that there aren't still the occasional snags. Getting high-performance hardware to work with Linux can still be tricky. "We have more hardware compatibility issues with Linux than with Windows, though it's getting closer," Gauvin said.

Though the emergence of enterprise builds such as SUSE and (particularly) Red Hat has led to a degree of standardisation, Linux remains much more diverse even within a tightly controlled environment such as the Hostworks datacentre.

"Whereas I can say with a fairly high degree of certainty that 90 per cent of our customers in the datacentre are on the same version of Windows, I wouldn't say that about Linux," Gauvin said.

"The other interesting flip side is that the TCO for Linux is slightly higher because the cost of the licences and the cost of the maintenance and the sort of effort we need to put in — it comes out a bit higher, though only a few per cent."

Can anything drive those costs down? Gauvin would be keen to see a more uniform certification system for Linux techs. "One of the things the Linux world in general could do better is lift their vendor-supported certifications, so we could start to compare skill sets."

Comments (7) | Email this

Share: Google | Facebook | del.icio.us | Digg | Reddit | Slashdot | StumbleUpon

• Why sex scandals are good for data management
• Unnatural language processing
• Crikey, Calvin, what were you thinking?
• What would Dr Who do?