Standards
As Stephen pointed out last week, and I touched on in reference to Force.com, there’s a fast emerging need for standards in cloud computing and we need them quick.
Keeping Things Running
As from-left-field example – monitoring and managing – while there are the beginnings of efforts, we don’t even have SNMP for the clouds. How about things orchestrating spinning up new instances and packaging?
These are just a few of the basic protocols and standards that keep on-premise things running. Glued on-top of the base level of monitoring and booting is a whole ‘nuther layer of configuration management, network performance management, dynamic scaling, and process to make sure you don’t shoot yourself in the foot.
What We Seem to Have So Far. No Much, But Something
Relative to that, perhaps the most interesting talk at Velocity was on EUCALYPTUS (see video above). Here, there’s a mix of standard and implementation. But, what’s key is the mixture of the neutral feel of the project along with the very real use-cases that managing a universities “cloud” comes with (pay attention to the access and identity management parts).
In a tea-leavers cross-over dream, Greg Papadopoulos (one of Sun’s CTO’s) mentioned EUCALYPTUS at Strucutre 08 as being interesting and one of the only real efforts around standards in cloud land.
The standards sentiment seems to have already spread, both in general and even to more narrow niches like data integration for the cloud. Indeed, I had several encouraging conversations last week about cloud standards that I’m hoping RedMonk can be a catalyst for moving along.
(We actually discussed cloud standards at length towards the end of our last podcast.)
Open Source?
When it comes to standards, the next question is of course around open source. Clearly, open source is beneficial for both developing and spreading standards: using running to code to get to de facto standards is dandy, and there’s nothing like free code to spread the use of a standard.
Interestingly, at all of the cloud conferences I went to last week, open source was very rarely mentioned, except as a sort of sneering thing. Indeed, sort of capturing the sentiment, I jotted down a stray quote from Structure’s VC panel about what cloud stuff can do, getting “beyond what you can do in open source.”
Of course open source is still valuable in exactly the same role is has now, in addition to standards support as: providing the middleware and libraries for software developers, both ISV and corporate. The question is what new definitions, benefits, and draw-backs it’d have in an open source world. Stephen poked the bear of re-defining what open source is a little while ago – the man has courage ;>
Disclaimer: Sun is a client, as is Hyperic.
Standards and open source accomplish different things. For instance, compare IEEE 802.3 (Ethernet) to Linux.
Ethernet is a structured and agreed upon way of accomplishing a specific task, the transmission of data. There's little or no wiggle room in the standard and as a result you're assured that just about any Ethernet product you purchase will work with any other. Hundreds of vendors have built Ethernet products over almost two decades, and even the oldest are still useable.
Linux, conversely, is more of an idea than a real standard. There are countless variations of Linux from SUSE to Red Hat, debian to Ubuntu. Most folks reading this blog have encountered applications that are tied to a specific implementation.
Open source is excellent at disseminating ideas and fostering experimentation. Indeed, most of the cloud computing services wouldn't exist without open source, including 3tera's own AppLogic. However, as we move to the need for interoperability, standards rather than open source will be required.
Howdy! This is kind of off topic but I need some advice from an established blog. Is it very hard to set up your own blog? I’m not very techincal but I can figure things out pretty quick. I’m thinking about making my own but I’m not sure where to begin. Do you have any ideas or suggestions? Cheers