Data Centers

PDC 2008: Making sense of Live Mesh

Justin James recently got the opportunity to speak with Microsoft's Jeff Hansen about Live Mesh. Find out why Justin says he's now more open minded about Live Mesh.


The big news from Microsoft's Professional Developers Conference 2008 (PDC 2008) was the launch of Windows Azure. There is a ton of confusion over what exactly Azure is -- an OS, a development platform, or something else? I had the chance to talk with Jeff Hansen, the General Manager of Microsoft's Live Services group after PDC 2008 wrapped up on October 28. He provided me with good information about this platform; he focused primarily on Live Mesh, which is his area of expertise.

In a nutshell, Live Mesh is a synchronization system for devices, clients, etc. In some senses, it is kind of like a USB hub. For example, you can take a picture with a Windows Mobile phone and push it to a printer to be printed or a digital picture frame to be displayed 300 miles away. But after learning it can do much more than that, I asked, "Would it be fair to say that this is the traditional client/server model paradigm extended beyond the LAN?" Jeff said, "Yes," and at that point, I felt like Live Mesh finally made sense to me after six months of not understanding it.

Last week, I discussed cloud computing in depth, and most of the reader feedback agreed that they would not be using it in the near future. A number of readers like the idea a lot and believe that it is how business will be done in the near future. I passed much of the skepticism to Jeff to see how he would respond.

Jeff told me that they see the "sweet spot" for Live Mesh is applications that will be accessed by many devices or instances of the application. When I asked about concerns regarding regulatory compliance (HIPAA, SOX) or data security, he said that the hurdles are neither technical nor legal but cultural. He agreed with me that cloud computing, SaaS, etc. does not have a high level of trust as an idea yet, but (like many of the comments I've read in these forums) he felt that over time, developers and businesses will feel comfortable putting more and more data outside of the LAN. He informed me that the BBC, Blockbuster, Travelport, and British Airways are all making use of Live Mesh or are planning to do so. Big name signups like this may be the best endorsement for the idea -- and not just Microsoft's offerings, but for its competitors as well.

In terms of security, the system is set up so that not even the people at Microsoft can read your data. While that has yet to be truly put to the test, it is good to know that they at least understand a major objection to cloud computing. Jeff and I also talked briefly about Microsoft's data centers. The company actually turns shipping containers into computing clusters and puts them in highly secured warehouses. This gives a high level of security, since it is fairly hard to walk out of a warehouse with a shipping container on a hand truck. At the same time, it allows Microsoft to have very consistent deployments and scatter data centers all over the place rather easily for redundancy. I have read a lot about these data centers over the last few months, and I believe that on the reliability and redundancy front, Microsoft has this nailed as well as can be expected, and certainly better than 99% of companies out there can. The security is still a question mark that I won't feel comfortable with until the system has had a real burn-in period. One thing I like is that the system is not "all or nothing" -- developers only put data in the cloud that they feel comfortable putting in the cloud.

Overall, I find the offering interesting. Compelling? Not yet. If I were writing code for a startup with no legacy code to maintain, and the data was not super sensitive, and the application needed to store data on servers (not on the client side), I would consider it. In other words, if I was writing a Flikr-type of site, I could see this being appealing. It's certainly more attractive than asking my IT department to pay people to add servers all day long just for storage. For other applications, I remain skeptical, but I am more open minded than I was before talking to Jeff.

Related TechRepublic resources


Disclosure of Justin's industry affiliations: Justin James has a working arrangement with Microsoft to write an article for MSDN Magazine. He also has a contract with Spiceworks to write product buying guides.


Get weekly development tips in your inbox Keep your developer skills sharp by signing up for TechRepublic's free Web Developer newsletter, delivered each Tuesday. Automatically subscribe today!


Justin James is the Lead Architect for Conigent.


I think I see myself doing something sooner, than later, in the cloud. More as a service provider, mind. I think it's still important to remember that users need something local. Maybe even the ability to print it. An example, most transport agencies have journey planners. Few seem to still have timetables and maps to go with them. You can go to Northern Ireland's Translink and get a plan to get from an obscure place in south Belfast to another obscure place in Enniskillen. But it takes a bit of wizardry to get maps and timetables for the routes. It's all very nice to have it online. What do you do when you're stuck somewhere where your bus connection doesn't work and no internet? I think the solution is to let the web application do its thing. It can work with the latest copies of all the data. Then give the user a chance to take and work with as much of that data as they need, locally.


"If I were writing code for a startup with no legacy code to maintain, and the data was not super sensitive, and the application needed to store data on servers (not on the client side), I would consider it." Right on. Cloud computing should be heralded for what it is, rather than for what it aims to be. It is a great idea that has real promise for large quantities of rapidly growing low-security data. It's not ready for prime time in a business culture that doesn't need outrageous accessibility and doesn't have such a thing as low-security data (I'm in the healthcare industry, and that's how we roll, anyway - you can do what you want with your data). Come on, let's be realistic - has Microsoft (or anyone, for that matter) made a large-scale system that is secure on the first run? For someone to pitch this kind of system as secure without mainstream exposure is naive at best. It will be a while before Live Mesh becomes industry standard, if it ever does. Still, like you, I'm probably more open to it now than I have been. I just don't think businesses have good reasons yet to move away from the traditional client-server model.

Justin James
Justin James

Are Live Mesh and Azure good ideas that are ahead of their time? Right on the money? Bad ideas? Solutions in search of problems? Let us know! J.Ja

Editor's Picks