Big Data, Cloud Storage and Security were Hot Topics at Cloud Camp Baltimore
Posted by Margaret Dawson on March 11, 2011
This week I attended another Cloud Camp, this one in Baltimore, Maryland. As has become the standard, the camp drew a strong crowd of some 150 mostly techies from the Baltimore and DC area, with a strong government contingency as you’d expect. Thanks to Booz Allen for hosting this event (Booz Allen is hiring by the way) and covering the bill for a nice meeting room and food/drink at the waterfront Marriott. (Although have to admit the upscale décor kind of clashed with the scrappy Cloud Camp culture, but the wine was good!!)
Hubspan was a sponsor once again, giving us a 5-minute lightening talk at the start of the camp, following Dave Nielsen’s definition of cloud computing. I spoke with my colleague and Hubspan software architect, Vipul Sabhaya, on how to approach the cloud and key pitfalls to avoid. Our presentation should be posted on the Cloud Camp site per request by attendees.
When it came to the un-panel portion of the camp, it was all about Big Data and how to manage it, store it, scale it, and secure it in the cloud. I’m guessing this was directly associated with the number of government agencies there, although there were companies like Comscore, who are leading the way with managing terabytes of data every day.
The real meat of Cloud Camps are the break-out sessions, which are created on the fly as attendees recommend topics that interest them or that leverage expertise in the room. We attended two sessions: Cloud Security and Scalability. The other aspect of Cloud Camp that I’ve seen repeatedly is the span of cloud knowledge in the room – from literally “what is the cloud?” to complex algorithms and advanced cloud architecture. This leads to interesting discussions.
In the Cloud Security break-out session, it was clear the majority of attendees (remember, a lot of government) still do not trust the public cloud for their most sensitive data or critical systems. In general, I think all attendees want to leverage the cloud, but there was a great deal of “we just have to have it behind our firewall”. Kind of makes me laugh, since this was the same “castle and moat” argument I used to hear when I was in the SSL VPN space, and IT guys were not willing to “punch holes” in their firewall, even though they were already there. There was discussion on ways to increase security in the cloud, such as leveraging strong key management and access control rules.
According to Vipul, the scalability session was the most interesting one of the night. The discussion was more about how to achieve close to linear scalability, and what people have done to get there. A lot of architectural discussions on the importance of “shared-nothing” data, and how to effectively shard your data, and how that is really the key in how much you can really scale. A lot of the NoSQL projects are aimed primarily around this concept. There was also some talk on the architecture of Google and Amazon and their scalability challenges, as well as how to architect an application such that it can also be linearly scaled.
Once again, hats off to Dave and the Cloud Camp crew. Another successful camping adventure.
Tags: Big Data, Booz Allen Hamilton, Cloud Camp, Cloud Computing, Cloud Security, Dave Neilsen
No Comments »
No comments yet.