How an earthquake affects hard drive and website performance

Yesterday’s 5.8-magnitude earthquake centered in Virginia was notable for three main reasons. First, earthquakes of this magnitude are incredibly rare on the east coast; second, it proved yet again that Twitter moves faster than seismic waves*; and finally, one of Amazon’s AWS (EC2/S3) data centers is based in north Virginia, about 60 miles north of the quake’s epicenter.

Amazon’s data centers are positively packed with hard drives — thousands of them — and every single one of them is incredibly sensitive to vibrations. Generally, though, the only time a rack of servers moves is during an earthquake — and when this happens, the read/write heads are instantly thrown out of position, and accelerometers then kick in to prevent data loss. As a result, it suddenly becomes very hard to access data on the disk, and with Amazon AWS this means that page load/response times ballooned. It is for this very reason that west coast (and Japanese) data centers usually use base isolation to earthquake-proof their racks.

The graph on the right comes from SeatGeek, which hosts its site with Amazon AWS.

Each of the vertical lines is two minutes — and so we can surmise that the earthquake was strong enough (and long enough) to affect the AWS hard drives for a couple of minutes. It’s also possible that the smaller spike on the left side of the graph was an early tremblor — but because the graph has been cropped (and its axes removed!) we don’t know if any aftershocks affected AWS.

Without an exact timeline it’s also impossible to say whether this spike was directly caused by the earthquake — but with a duration of just two minutes, it’s unlikely that it was because of increased traffic; a burst of traffic would last longer, as everyone and their mother starts hitting Twitter, Facebook, and news websites. On the other hand, it’s safe to assume that this same spike would be felt by every east coast data center — but lots of anecdotal evidence suggests that the web itself didn’t seem to slow down during the quake. Surely, if almost every hard drive in and around Washington DC was taking 10 times as long to access data, we would’ve noticed it?

Perhaps everyone was simply too busy tweeting to notice the slowdown, though. Last month, an ExtremeTech reader left a comment with a link to an amazing video (embedded below) that all but confirms that yesterday’s earthquake crippled Amazon AWS. In the video, a charismatic Australian (Brendan Gregg) shouts into a rack of hard drives, and the vibrations caused by his only-slightly-louder-than-ambient hollering is enough to slow down disk I/O significantly.

In the accompanying blog post, Gregg also notes that even a missing screw in a mounting bracket can introduce enough vibration to cause intermittent spikes in access time — so if you’re a serious gamer, remember to make sure your hard drives are mounted properly with anti-vibration grommets, and if you have a few thousand spare dollars, invest in one of those base isolation systems for your desktop rig, too. Alternatively, you could always replace your hard disk with a solid-state drive; moving parts are so passé, after all.

see more.....How an earthquake affects hard drive and website performance

IBM builds 120 petabyte cluster out of 200,000 hard drives

Smashing all known records by a multiple of 10, IBM Research Almaden, California, has developed hardware and software technologies that will allow it to strap together 200,000 hard drives to create a single storage cluster of 120 petabytes — or 120 million gigabytes. The drive collective, when it is complete, is expected to store one trillion files — or to put it in Apple terms, two billion hours of MP3 music.

The data repository, which currently has no name, is being developed for an unnamed customer — but with a capacity of 120PB, its most likely use will be the storage device for a government-owned or federally-funded supercomputer or other high-performance computing (HPC) application; 120PB is the kind of capacity that you need to store global weather models or infinitely detailed weapon system simulations, both of which are rarely carried out by commercial interests. Alternatively, it could be used to store a large portion of the internet (or data about its users) for Google or Facebook, or another client with very deep pockets. The largest systems currently in existence are generally around 15 petabytes — though, as of 2010, Facebook had a 21PB Hadoop cluster, and by now it’s probably significantly larger.

Exact details about the software and hardware isn’t given by IBM, but we do know that it features a new-and-updated version of IBM’s General Parallel File System (GPFS). GPFS is a volume-spanning file system which stores individual files across multiple disks — in other words, instead of reading a multi-terabyte high-resolution model at 100MB/sec from a single drive, the same file can be read in a massively parallel fashion from multiple disks. The end result is read/write speeds in the region of several terabytes per second — and, as a corollary, the ability to create more than 30,000 files per second. GPFS also supports redundancy and fault tolerance: when a drive dies, its contents are rebuilt on a replacement drive automatically by the governing computer.

On the hard drive side of things, if we divide 120PB by 200,000 you get 630GB — and once you factor in redundancy, it’s fairly safe to assume that the drives are all 1TB in size. We also know that every single one of the 200,000 drives will be watercooled with presumably the largest and most complicated bit of plumbing ever attempted — but considering IBM’s penchant for watercooling its top-end servers, that’s hardly surprising (though we still hope to post a photo of the system once it’s complete).

As it stands, supercomputers — and large-scale science experiments like the LHC — can produce (and compute) far more data than can be feasibly stored. IBM’s system, it is hoped, will be a data repository that goes some way to bridge the gulf between silicon, Moore’s law-governed technology — and electro-mechanical storage. Of course, on the other hand, perhaps it’s time to stop playing around with hard drives and start building mass storage arrays out of flash memory…

Read more at Technology Review, or about IBM Research Almaden or GPFS.

The New " Mac iStack "

Is the PC dying? You’d be forgiven for thinking it, what with the largest PC maker bowing out, and raving iPad-toting Jobsians pointing at the vast sales of smartphones, tablets, and other post-PC devices.

It doesn’t really matter if the desktop PC is dying, however, if there is nothing to replace it. Tablets and smartphones might be nibbling at the PC’s heels, but they certainly won’t replace them any time soon. Likewise, MacBook Pros and Ultrabooks are certainly forging towards desktop-like processing power in a mobile form factor, but heat displacement issues and lack of upgradability mean that they will never truly usurp desktop PCs.

In spite of all this, though, it’s impossible to deny that the desktop tower has run its course. Their gargantuan footprint, their oversized power supplies, their excessive number of expansion slots and drive bays — the truth is, almost every household could now make do with a desktop PC the size of the Mac Mini, as long as it has plenty of external connectivity. If it wasn’t for the power users, in fact — if graphic designers, video editors, and gamers weren’t so darn needy — then the big beige box would’ve vanished years ago.

What if Steve Jobs could craft a magical device that fulfills the needs and desires of Apple zealots and earnest power users alike, though?

There is a rumor going around that Apple will introduce an entirely new product line before the end of the year, you see. Combine this rumor with the fact that Apple’s monstrous (41-pound!) aging aluminium cheese grater, the Mac Pro, is unwieldy, incredibly expensive, and hasn’t yet been updated to Intel Core i processors, and you have the possibility that Apple’s new product might actually replace the Mac Pro.

Now… what if the replacement is a new, modular Mac system. A stackable system, using the Mac Mini form factor, and sewn together with Thunderbolt — or perhaps, in a year or two, with the much higher-bandwidth-but-shorter-distance fiber optic Light Peak interconnect.

You can see Jobs now, bouncing around a polished obsidian keynote stage, in an ebony cashmere sweater, and holding aloft a glistening white box that he heralds as the centerpiece of the onrushing post-PC world. It would be called the Mac Desktop, and it would combine to form [the] iStack.

This isn’t as crazy as it sounds. If you’re a normal home user who just wants to word process, listen to music, and watch TV, buy one Mac Desktop; you get all the power of your MacBook Pro or Air, but for half the price and in a tidier form factor that you can hide away in a cupboard somewhere. If you’re a gamer, just buy two. If you need some serious processing power for compilation, rendering, or audiovisual editing, strap four Mac Desktops together and make a bonafide iStack. Thunderbolt (a glorified PCIe bus) and Apple’s Grand Central Dispatch symmetric multiprocessing (SMP) interface would make short work of connecting them all together, too.

It gets better: if you’re worried about graphics capabilities, don’t forget that the new iMac is basically a Mac Mini and Thunderbolt Display strapped together. It isn’t equipped with a top-level graphics card at the moment — an AMD Radeon HD 6970M — but that isn’t to say that it couldn’t be outfitted with something gamerific. Connecting your stack of MacTops to your iMac is just a matter of running a Thunderbolt cable — and again, the stack can be kept in a cupboard because Thunderbolt allows you to connect every peripheral to the iMac.

But why stop there? The Apple Time Capsule has a similar form factor to the Mac Mini; let’s put one of those in the iStack. And an Apple TV box, while you’re at it. Before you know it, you have a single, pearly stack that’s capable of doing just about anything. As a matter of course, your iPhone and iPad could communicate with the whole caboodle via wi-fi. Imagine if a future version of iOS let you use your mobile device as a wireless display…

Finally, underpinning it all is Apple’s new iCloud service. You will zap a QR code on a poster with your smartphone, and the movie or album will automatically download to your iStack at home. The photos and videos that you shoot will be automatically backed up the iCloud, but also your iStack Time Capsule at home. You will be able to play a game on the way to work, and then pick up from where you left off on your Apple TV.

In short, iStack will be everything that today’s PC wants to be.

Need For Speed World, (free online NFS)

What is Need for Speed World?

Need for Speed World is the FREE to play online racing game where you decide how to play in a massively multiplayer world. Connect with friends or race against the world as you rise through the rankings. Race the very latest cars or relive classic NFS moments with the ever-expanding car list. Stamp your identity on the cars with an endless combination of paints and liveries. Play for Free! Play it Now!

What is Play4Free?

With Play4Free games you don’t pay anything to download and play the game… it’s completely free to play. You decide how much you want to pay… if anything. If you want to spend money you can; buy new cars or customize the performance of your current car but if you don’t want to spend any money there’s nothing to stop you racing your way to the top of the leaderboards; racing the fastest cars all the way to level 50.

download : NFS World

Moonbase Alpha PC Game Overview

NASA has once again landed on the lunar surface with the goal of colonization, research, and further exploration. Shortly after the return to the Moon, NASA has established a small outpost on the south pole of the moon called Moonbase Alpha. Utilizing solar energy and regolith processing, the moonbase has become self-sufficient and plans for further expansion are underway.

Moonbase Alpha Game ScreenShot -- Repairing the Life Support System

In Moonbase Alpha, you assume the exciting role of an astronaut working to further human expansion and research. Returning from a research expedition, you witness a meteorite impact that cripples the life support capability of the settlement. With precious minutes ticking away, you and your team must repair and replace equipment in order to restore the oxygen production to the settlement.

Team coordination along with the proper use and allocation of your available resources (player controlled robots, rovers, repair tools, etc.) are key to your overall success. There are several ways in which you can successfully restore the life support system of the lunar base, but since you are scored on the time spent to complete the task, you have to work effectively as a team, learn from decisions made in previous gaming sessions, and make intelligence decisions in order to top the leaderboards.

download : Moonbase Alpha

Facebook-Integrated Skype v5.5

Just as soon as Facebook and Skype announced their major integration deal a few weeks back, Skype for Windows v5.5 Beta hit the download page. Now, the final version is here: Skype 5.5 for Windows is available for download, and based on user feedback since the June beta, there have been a number of updates that the company says warrant an upgrade.

The Skype 5.5 for Windows update comes with a deeper Facebook integration. Without having to leave Skype, you can check out when your Facebook friends are online and instant message with them from the new Facebook Contacts Tab (located in between the Skype and Recent Contacts tabs). You can also view and update your Facebook status and even view comments or ‘like’ posts from your Facebook Wall, directly from Skype Home.

They also improved the controls for video and group video calls for Windows, enhanced video call reliability and made several design changes to improve overall experience. Plus for all the emoticon fanatics, the Skype emoticons have received an update, with smoother animations that stand out.

Download : www.skype.com

Windows 8 in April 2012? Could Be!

ZDnet’s Mary Jo Foley is reporting on a rumor:Microsoft may be trying to finish up work on Windows 8 by April of next year. She thinks it’s plausible–or at least not obviously crazy. Me, too. For one thing, the conventional wisdom that the OS is likely to show up for the holiday 2012 season is, as far as I know, based more on history than on anyone knowing anything specific about Windows 8. For another, Microsoft has a huge incentive to get this thing out the door–not so much for its PC business, but for tablets, where it’s not yet really in the game and won’t be until Windows 8 is available. And Steven Sinofksy, the Microsoft exec in charge of Windows, has a pretty good track record for exceeding expectations when it comes to shipping products in a timely fashion. (Enough so that I think that anyone who parrots the classic “Microsoft never gets anything out the door” meme hasn’t been paying attention over the past few years.)

We should have a better sense of when Windows 8 is likely to show up once Microsoft holds its BUILD conference in September. That’s when the company will probably release some sort of preview or beta edition of the new OS and talk a little more about its timetable, without revealing all.

Me, I vote for Microsoft taking its own sweet time. A bunch of companies have already proven that it’s a bad idea to rush out devices based on tablet operating systems that aren’t fully cooked yet. And with the possible exception of Samsung, no company has established itself as a serious ongoing competitor to the iPad. Which means that a wonderful Windows 8 that appears late in 2012 might have a better chance of being a big deal than a so-so one that makes a surprisingly early debut.

How to work with Google Docs

Google Docs is a free, Web-based office suite, and data storage service offered by Google. It allows users to create and edit documents online while collaborating in real-time with other users. Google Docs combines the features of Writely and Spreadsheets with a presentation program incorporating technology designed by Tonic Systems. Data storage of files up to 1 GB each in size was introduced on January 13, 2010, documents created inside Google Docs not counting towards this quota.

For those unfamiliar, Google Docs provides a way to create and edit documents from the cloud. That means that you can do a lot of typical word processing, spreadsheet work, and presentation creation without installing any bulky software on your computer. That also means that as long as you can remember your Google password, you can retrieve your docs from any computer connected to the internet; the same is true for anyone else that you invite to collaborate with you.

What you need to access Google Docs?

Open the Google Docs website (see Resources below). When the Google login page appears, simply enter your google Id and pass word box on the right. If you don’t have a Google account, create a free account, which will takes only seconds.

Create your first document by clicking on the “New Document” button at the top of the page, then choosing the kind of file you want to create from three options–”Document,” “Spreadsheet” or “Presentations.”

Begin using your document just as you would a Microsoft Word, Excel or PowerPoint document. You will see pretty much the same task bars as you see while working on any of those Microsoft programs. For example, if you’re working on a text document, you’ll see buttons for fonts, bold, italics, underlining, indents, alignment, cutting, copying, pasting, and so on.

Save your document by clicking on the “File” button and choosing “Save” in the drop-down menu. Rename your document by clicking on “File” and then “Rename.” Once you save or rename a document, you will be prompted for a file name. Choose a file name and click “OK,” and you’re done.

Choose tags for your documents. Because Google doesn’t use folders to hold your documents, you will need a way to retrieve them, especially if your list of documents becomes lengthy. By tagging as soon as you complete and name a document, you save valuable time in the long run. To tag an article, choose it from your list of files by clicking the check box next to it. Then click on “Tags” at the top of the page, and in the drop-down menu, you will see tags for your document suggested by Google. Choose from those and/or add your own tags.

Share a document with others by opening the Google Docs page. When you see your list of documents, click on the “Share Now” button to the right of your document title. You will be prompted to enter the email addresses of people with whom you want to share the document. Enter the emails and decide whether or not you want these individuals to be “Collaborators” or “Viewers.” Viewers can only read your document, and collaborators have your permission to edit the document. Anyone who wants to view the document has to sign up with Google.

Publish your Google document to the Web by opening the document from your list then clicking on “Publish,” which is located in a tab at the top and the far right. Print your document by choosing “File” in your task bar, then “Print.” If you would like to print the document from a PDF, choose “File” and then “Save as PDF” from the drop-down menu. Once Adobe Acrobat has converted your document into a PDF, you can print it directly from Google Docs.

New Control Over Privacy on Facebook

Privacy worries have bedeviled Facebook since its early days, from the introduction of the endless scroll of data known as the news feed to, most recently, the use of facial recognition technology to identify people in photographs.

At the nub of all those worries, of course, is how much people share on Facebook, with whom and — perhaps most important — how well they understand the potential consequences.

The company has struggled to find a balance between giving users too little control over privacy and giving them too much, for fear they won’t share much at all. Seeking a happy medium, Facebook announced changes on Tuesday that it says will help users get a grip on what they share.

When the changes are introduced on Thursday, every time Facebook users add a picture, comment or any other content to their profile pages, they can specify who can see it: all of their so-called Facebook friends, a specific group of friends, or everyone who has access to the Internet. These will be indicated by icons that replace the current, more complicated padlock menu.

Similar controls will apply to information like users’ phone numbers and hometowns and whether they like, say, death metal bands, on their profile pages. Users will no longer have to seek out a separate privacy page to tweak who sees how much of that personal information. Nor will they have to bother to remember what those settings were.

Company officials say they hope the changes will simplify the process of establishing who knows what about your life on the Internet — and hopefully, save a few people the embarrassment of unwittingly sharing too much.

“We want to make this stuff unmistakably clear,” Chris Cox, vice president for product at Facebook, said in an interview. “It has to be clear that Facebook is a leader in how people control who sees what.”

Implicit in these changes is the challenge brought on by Facebook’s own success. It is used by 750 million people worldwide, with varying degrees of knowledge about what it means to have a life online. There is the looming prospect that the company will go public, along with the abiding concern about potential government regulation or litigation stemming from privacy issues.

Not least, there is the need for Facebook to cultivate the trust of its users, amid growing competition from Google’s nascent social networking service, Google Plus, which emphasizes more compartmentalized communications with different sets of friends and acquaintances.

Facebook dismissed the notion that the changes were fueled by competition. Company officials took pains to tell reporters that they had briefed privacy advocates on the changes — including those who have been critical of Facebook — and solicited their feedback.

It is too early to tell whether users will find the changes more inviting or simpler, or whether they will reduce what Kurt Opsahl, a lawyer with the Electronic Frontier Foundation, called “the amount of unintentional oversharing on Facebook.”

Privacy advocates warned that the new tools did not address a concern about sharing location. One Facebook user can publish information about another user’s whereabouts without his or her consent — whether it’s an employee at the beach on the day he or she called in sick or a husband at a strip club without his wife’s knowledge.

Other privacy experts say that if users believe they have control over who sees what, they are more likely to share.

“I think it’s part of an evolution to push back at the notion that Facebook is trying to trick you into sharing,” said Jules Polonetsky of the Future of Privacy Forum, which is based in Washington. “You’re more likely to do so when you know what you’re doing.”

The new tools represent a departure from Facebook’s more recent approach, in which users found much of what they posted — tags, photos and so on — to be widely accessible unless they explicitly specified otherwise. The default position, in other words, was to opt for sharing.

Mr. Cox said the new tools were meant to demystify privacy controls and ensure that Facebook users were never “surprised” by what others could see about them.

That includes pictures in which they have been “tagged.” No longer will an unflattering or compromising photograph appear on your profile page without your consent, though the publisher of the photograph can still keep it up on his or her own page. Users will be able to approve every picture in which they are tagged before it appears on their profile pages.

Additionally, the privacy option that is now called “everyone” will instead be called “public.” Facebook executives say they want to dispel any doubts about what the setting means. If you click “public,” that means anyone who is online can see what you are posting, including perfect strangers — or, worse, parents, prospective employers and your ex-wife’s divorce lawyers.

“We need to offer fine granularity in order to be a universally usable tool,” Mr. Cox said. With the new settings, “it’s more visual and prominent who the audience is.”

Indeed, company officials say feedback from users suggest that pictures work better than words. So now, icons guide the way. “Public” is represented by a globe; “friends” by a pair of heads.
A version of this article appeared in print on August 24, 2011, on page B1 of the New York edition with the headline: New Control Over Privacy On Facebook.