Hello after a long time now in Technology section a new and great news announcing. In near future data transfer on large scale will be very easy and it will flexible to transfer data from multi places at a time to another place which may be your destination for data sending,
Now we are talking about Cloud data computing that offer to users offload their number crunching problems to powerful servers elsewhere, but here normally we facing problem that large files of data is hard to transfer and it's take large time to upload. In fact, it can sometimes be quicker and cheaper to copy your files to a hard drive and stick it in the post.
That's why two computer scientists at the University of Illinois at Urbana-Champaign are developing a system to calculate the most cost-efficient way to transfer data, using a mixture of the internet and postal services.
This new system, called Pandora, aims to minimise data transfer times for a given budget. Imagine you've got 1.2 terabytes (TB) of data at one location and 0.8TB at another - two business address at opposite ends of the country for example - and you want to upload both sets of data to Amazon's EC2 cloud computing service. You could just upload them to the Amazon's server over the internet, but Amazon charges $100 per TB, so the total cost would be $200. Uploading that much data also takes a while - it could be weeks, depending on your available bandwidth.
It turns out the cheapest approach is to combine the two sets of data over the internet, then download it onto a 2TB disk and ship it to Amazon, which charges a handling fee of $80 per disk and a much lower rate per terabyte for uploading data from the disk. Although you save money - the total cost would be roughly $120 - this option is slower than simple uploading via the internet, taking 20 days.
A faster but more expensive option is to post a 2TB disk between your two locations, combine the data on a hard disk, then send it on to Amazon to process.
Phew. Even with just two sets of data at different locations, its clear that unravelling the various options is a complex problem - what happens when you've got hundreds? The researchers hope that Pandora can take the headache out of data transfer by finding the best solutions.
The tool is mostly intended for researchers and businesses that regularly need to shuffle their data around, but home users might also start to need it as we store more and more data in the cloud.
Now we are talking about Cloud data computing that offer to users offload their number crunching problems to powerful servers elsewhere, but here normally we facing problem that large files of data is hard to transfer and it's take large time to upload. In fact, it can sometimes be quicker and cheaper to copy your files to a hard drive and stick it in the post.
That's why two computer scientists at the University of Illinois at Urbana-Champaign are developing a system to calculate the most cost-efficient way to transfer data, using a mixture of the internet and postal services.
This new system, called Pandora, aims to minimise data transfer times for a given budget. Imagine you've got 1.2 terabytes (TB) of data at one location and 0.8TB at another - two business address at opposite ends of the country for example - and you want to upload both sets of data to Amazon's EC2 cloud computing service. You could just upload them to the Amazon's server over the internet, but Amazon charges $100 per TB, so the total cost would be $200. Uploading that much data also takes a while - it could be weeks, depending on your available bandwidth.
It turns out the cheapest approach is to combine the two sets of data over the internet, then download it onto a 2TB disk and ship it to Amazon, which charges a handling fee of $80 per disk and a much lower rate per terabyte for uploading data from the disk. Although you save money - the total cost would be roughly $120 - this option is slower than simple uploading via the internet, taking 20 days.
A faster but more expensive option is to post a 2TB disk between your two locations, combine the data on a hard disk, then send it on to Amazon to process.
Phew. Even with just two sets of data at different locations, its clear that unravelling the various options is a complex problem - what happens when you've got hundreds? The researchers hope that Pandora can take the headache out of data transfer by finding the best solutions.
The tool is mostly intended for researchers and businesses that regularly need to shuffle their data around, but home users might also start to need it as we store more and more data in the cloud.
Post a Comment