Web servers that store data locally save time on database searches but sometimes serve up obsolete results. A new system solves that problem.

 

 

[gallery]

Today, visiting almost any major website — checking your Facebook news feed, looking for books on Amazon, bidding for merchandise on eBay — involves querying a database. But the databases that these sites maintain are enormous, and searching them anew every time a new user logs on would be painfully time consuming. To serve up data in a timely fashion, most big sites use a technique called caching. Their servers keep local copies of their most frequently accessed data, which they can send to users without searching the database.

But caching has an obvious problem: If any of the data in the database changes, the cached copies have to change too; moreover, any cached data that are in any way dependent on the changed data also have to change. Tracking such data dependencies is a nightmare for programmers, but even when they do their jobs well, problems can arise. For instance, says Dan Ports, a graduate student in the Computer Science and Artificial Intelligence Lab, suppose that someone is bidding on an item on eBay. The names of the bidders could be cached in one place, the value of their bids in another. Making a new bid updates the database, but as that update propagates through the network of servers, it could reach the value cache before it reaches the name cache. The bidder would see someone else’s name next to her bid and think she’d been beaten to the punch. “They might see their own bid attributed to somebody else,” Ports says, “and wind up in a bidding war with themselves.”

MIT researchers have developed a new caching system that eliminates this type of asymmetric data retrieval while also making database caches much easier to program. Led by Ports and his thesis advisor, Institute Professor Barbara Liskov, who won the 2008 Turing Award, the highest award in computer science, the research also involves associate professor Sam Madden, PhD student Austin Clements, and former master’s student Irene Zhang. Ports presented the system on Oct. 5 at the USENIX Symposium on Operating Systems Design and Implementation in Vancouver.

Transact locally

Unlike existing database caching systems, Ports and Liskov’s can handle what computer scientists call transactions. A transaction is a set of computations that are treated as a block: None of them will be performed unless all of them are performed. “Suppose that you’re making a plane reservation, and it has two legs,” says Liskov. “You’re not interested in getting one of them and not the other. If you run this as a transaction, then the underlying system will guarantee that you get either both of them or neither of them. And it does this regardless of whether there are other concurrent accesses, or other users are trying to get seats on those flights, or there are machine failures, and so forth. Transactions are a well-understood technique in computer science to achieve this kind of functionality.” Indeed, it’s the idea of transactions that gives the new system its name: TxCache, where “Tx” is a shorthand for “transaction.”

TxCache also makes it easier for programmers to manage caches. “Existing caches have the approach that they just make this cache and tell the programmer, ‘Here’s a cache: You can put stuff in it if you want; you can get stuff out of it if you want,’” says Ports. “But figuring out how to do that is entirely up to you.” TxCache, however, recognizes that a computer program already implicitly defines the relationships between stored data. For instance, a line of code might say that Z = X + Y, which is an instruction to look up X, look up Y, and store their sum as Z. With TxCache, the programmer would simply specify that that line of code — Z = X + Y — should be cached, and the system would automatically ensure that, whenever any one of those variables changed, the cached copies of the other two would be updated, everywhere. And, of course, it can perform the same type of maintenance with more complicated data dependencies, represented by more complicated functions.

Bean counting

According to Liskov, the key to getting TxCache to work was “a lot of bookkeeping.” The system has to track what data are cached where, and which data depend on each other. Indeed, Liskov says, it was the fear that that bookkeeping would chew up too many computing cycles that dissuaded the designers of existing caching systems from supporting transactions. But, she explains, updating the caches is necessary only when data in the database change. Modifying the data is a labor-intensive operation; the bookkeeping steps are comparatively simple. “Yes, we are doing more work, but proportionally it’s very small,” Liskov says. “It’s on the order of 5 to 7 percent.” In the researchers’ experiments, websites were more than five times as fast when running TxCache as they were without it.

“The trouble with large-scale services like Bing and Amazon and Google and the like is that they operate at such a high level of scalability,” says Solom Heddaya, a partner at Microsoft and infrastructure architect for Bing, Microsoft’s search engine. “On a single request from the user searching for something, there are many, many applications that get invoked in real time, and they together will use tens of thousands of servers.” On that scale, Heddaya says, some kind of caching system is necessary. But, he says, “until this paper came along, people building these systems said, ‘Hey, we will shift the burden to the programmer of the application. We will give you the convenience of caching, so that we bring the data closer to where the computation is, but we will make you worry about whether the cache has the right data.”

Heddaya cautions that, unlike some other caching systems, the MIT researchers’ offers significant performance improvements only for sites where reading operations — looking up data in the database — greatly outnumber writing operations — updating data in the databases. But according to Ports, “Adding support for using caching during read/write transactions is one of the things we're thinking about now. There aren't any major technical obstacles to doing so: It's mainly a question of how we can do so without introducing unexpected effects that make life more difficult for users and programmers.”

Nokia expands Symbian^3 family with Nokia E7, Nokia C6-01 and Nokia C7.

Symbian^3 family of devices is growing with the announcement of three new products across Cseries and Eseries. Joining the Symbian^3 based Nokia N8 are the all new touchscreen and QWERTY business device, the Nokia E7, a compact touchscreen C6 and the sleek touchscreen Nokia C7.

FEATURES OF NOKIA E7:

1.Billed as the ultimate business smartphone, the Nokia E7 boasts Microsoft Exchange ActiveSync support alongside a luscious 4-inch touchscreen display, which uses Nokia’s ClearBlack technology for improved outdoor visibility.

2.The Nokia E7 also comes with a slide out four-row QWERTY keyboard making it the ideal mobile business device. Despite the addition of a full keyboard, the E7 is only a smidge deeper than the Nokia N8 at 13.6mm thin.

3. The device also boasts an 8-megapixel snapper and comes complete with 16GB of on-board storage.

4..As a business device, the Nokia E7 comes sporting a raft of security features including business grade device lock and wipe functionality, secure intranet access and remote device management.

5. These come alongside its connectivity and productivity prowess, with support for Mail for Exchange, Microsoft Communicator Mobile and Microsoft Share Point Server.

6. Business folk on the move will also benefit from free walk and drive navigation thanks to Ovi Maps including premium guides from the likes of Lonely Planet and Via Michelin.

7.Like the Nokia N8, the Nokia E7 also boasts an HDMI connection and Dolby Digital Plus Surround Sound. With the 8-megapixel camera, users will be able to shoot and edit HD quality images (including 720p video) and sound.

8. As a Symbian^3 device, users can customise up to three homescreens and get live updates from the likes of Facebook and Twitter live on their homescreens.

9.Those on the move will welcome the Nokia E7′s travel-friendly 18 days of standby time and up to 9 hours of talktime. The device, which will be available in dark grey, silver white, green, blue and orange will be available in the last quarter of 2010 and is expected to retail for €495 before taxes and subsidies.

Joining the Nokia E7 and Nokia N8 as part of the Symbian^3 family are the all-new Nokia C6 and Nokia C7. Both of these devices raise the bar for design and social networking amongst their already good-looking Cseries brethren. Homescreen access to the latest updates from your favourite social networks are offered on tap for both devices.

First in the new Cseries lineup is the sleek C7 which boasts a 3.5-inch AMOLED display and is a hand-friendly 10.5mm thin. Like the E7 this device also sports an 8-megapixel camera with 8GB of onboard storage (expandable to 40GB with microSD).

The Nokia C7 will also shoot HD video in 720p and comes with free global walk and drive navigation and a digital compass. City users will welcome the active noise cancellation for clearer voice calls whilst the battery will keep calls going for up to 9.5 hours (in GSM mode) and stay on standby for an astounding 27 days (3G mode). The Nokia C7 is expected to go on sale before the end of 2010 and will be available in black, metal or brown for an estimated retail price of €335 before taxes and subsidies.....



The Symbian^3 family of products grow to four, all to be available before the end 2010. This latest version of Symbian is packing 250 new features and improvements whilst remaining familiar to the millions of existing Symbian users worldwide. With multiple homescreen support, multitasking and gesture-based interactions, this latest version of Symbian^3 is a big leap forward.

We’re excited to see the Symbian^3 family of devices come to life and we can’t wait to get our hands on them after the keynotes have finished. Stay tuned for more later, meanwhile let us know what you think in the comments below

Firefox 4 ,with sync and panorama ,provides a faster and much more personalized browsing experience.

New look

Firefox 4 beta comes in a sleeker and easier to use look with a firefox menu button on top left and new location of tabs.The tabs have been given top visual priority.

SYNC

Found an interesting website just when you were about to leave??Browser is now mobile!Using sync you can access your saved passwords, history ,open tabs and all other customizations using any computer or mobile device that uses firefox.For this you need to create a Firefox Sync account.

For instructions on how to use sync view  http://videos.mozilla.org/serv/firefox4beta/syncvideo5final640x3%23161DA3.ogv

PANORAMA



Firefox introduces panorama to organize tabs.You can drag tabs into tabspaces to make groups or piles and then open or expand respectively to navigate through them.Groups can be named, expanded or squeezed and managed quite visually.This helps to keep distractions out of sight for ‘clutter-free’ browsing.

For instructions on how to use panorama view

http://videos.mozilla.org/serv/firefox4beta/grouptabs.theora.ogv

Feedback

Using the Feedback add-on on the upper right corner you can give feedback about the browser any time.

Feedback add-on privacy policy

http://www.mozilla.com/en-US/firefox/beta/feedbackprivacypolicy/

Managing plugins

To ensure uninterrupted browsing ,firefox provides support in case commonly-used plugins Adobe Falsh, Quicktime or Microsoft Silverlight crash or freeze.You can simply reload the page and restart the plugin.

Others

-add-ons can be managed better and can be installed without restarting the browser.

-supports the webM format so that you can watch Hd videos.

-cleaner graphics

-super fast scrolling on complex web pages.

Microsoft releases the beta version of Morro antivirus or free security essentials as a move to provide integrated security and get a hold on vast ever growing anti-virus market.
The key features of this antivirus is that update and upgrades are done automatically, its very easy to install and easier to use It gives  you get high-quality protection against viruses and spyware, including Trojans, worms and other malicious software. And above all, its avaliable free of cost to all user having geniune windows.
it is light on your system and dont use a lot of resources and above all there are no annoying alerts