Information Technology News.

MIT creates method to shadow packets as they pass through networks

Share on Twitter.

Sponsered ad: Get a Linux Enterprise server with 92 Gigs of RAM, 16 CPUs and 8 TB of storage at our liquidation sale. Only one left in stock.

Sponsered ad: Order the best SMTP service for your business. Guaranteed or your money back.

March 23, 2017

A student at the Massachusetts Institute of Technology has created a new method at shadowing various data packets as they pass through today's busy networks, in an effort to help take the risk out of experimenting with various modifications to network configurations as they are being made in real time.

Such old stuff as the usual disaggregation of the control plane from the data plane has been around for ever and experienced network admins have found their way around those problems a long time ago.

However, according to MIT graduate student Amy Ousterhout and her collaborators, there's good reason to think that another level of disaggregation might be useful for their daily jobs of making networks do what they do best: send data over short or long distances on the internet as well as on private networks.

To be sure, 'Flexplane' takes the software routing-- the control plane, in SDN-speak and splits it in two-- there's the real network, where packets move, and there's also an emulated network that's a software model. It sounds complicated but it isn't.

“The emulator maintains a model of the network topology in software, and users implement their various schemes in the emulated network routers. Before each packet is sent on the real network, a corresponding emulated packet traverses the emulated network and experiences behavior specified by the resource management scheme. Once this is complete, the real packet traverses the network without delay”, the whitepaper asserts.

But why would a network or system admin go though all of that? As Ousterhout explains, it means that network administrators can experiment with changes to the control plane and get the emulator to tell them what's going to happen.

They can then only push new configuration settings into production if it's not going to break things. It's a much safer and faster way of doing things, while still making sure that all will be okay.

The whitepaper asserts that Flexplane can still keep up with a line rate throughput up to 760 Gbps on a 10-core server, which is a lot faster than a full software implementation of a router, and fast enough to operate a lot of production environments, even though it's targeted at academic and experimental users, for now at least.

'Flexplane' gets that performance because unlike a software router, its emulator doesn't try and deal with whole packets. Instead, the header fields of an ingress packet is all that's handed to the emulator. It's a clever way of doing things, one observer said.

“The way it works is, when an endpoint wants to send a data packet, it first sends a request to this centralised emulator,” Ousterhout explains in her MIT release.

“The emulator does in software the exact scheme that you want to experiment with in your own network. Then it tells the endpoint when to send the TCP or even UDP packet so that it will arrive at its destination as though it had traversed a network running the programmed scheme,” asserted Ousterhout.

As well as allowing a network administrator to set up and test some config modifications before they go to production, there's another more academic application of the system: protocol testing.

For example, MIT's media release system uses the industry-standard Transmission Control Protocol (TCP) as its main system. It can be extremely difficult to replace, partly because who's going to set up a hyperscale data centre as the test case for a new protocol? Flexplane can solve that.

Reconfigurable hardware such as using FPGAs is very labor intensive, and software routers can't reach high enough throughput to provide a realistic simulation under similar circumstances, she adds.

MIT quotes a Yale associate professor, Minlan Yu, in explaining why the emulation is a useful approach: “Many potential technical issues can happen at the interactions between applications and the network stack, which are difficult to understand and troubleshoot by simply simulating the traffic.”

Ousterhout, a graduate student in electrical engineering and computer science, worked with another graduate student Jonathan Perry, advisor Hari Balakrishnan, the Fujitsu Professor in Electrical Engineering and Computer Science, and Facebook network engineer Petr Lapukhov.

Flexplane will get its first public testing next week at the Usenix Symposium on Networked Systems Design and Implementation. The official Conference in Vancouver, BC is for August 16–18, 2017, but next week will be the first time the public will have a chance to have a look at it hands-on.

Source: The Massachusetts Institute of Technology, Boston, MA.

Sponsered ad: Get a Linux Enterprise server with 92 Gigs of RAM, 16 CPUs and 8 TB of storage at our liquidation sale. Only one left in stock.

Sponsered ad: Order the best SMTP service for your business. Guaranteed or your money back.

Share on Twitter.

IT News Archives | Site Search | Advertise on IT Direction | Contact | Home

All logos, trade marks or service marks on this site are the property of their respective owners.

Sponsored by Sure Mail™, Avantex and
by Montreal Server Colocation.

       © IT Direction. All rights reserved.