Last week I was in the bay area for 4 conferences (Force.com, Velocity, CloudCamp, and Structure) centered on cloud computing along with, of course, plenty of meetings and bar-side chats with the good folks running around there.
Cloud Conference Series – Five Parts to Cloud Madness!
In addition to this first part, there’s a handful of notes on different aspects of the week and, more interestingly, overall screeding on cloud computing:
- Part 1 – Velocity, Cloud Camp, Structure (you’re soaking in it!)
- Part 2 – You Think It’s Bad Now: Get Ready for Cloud-* – Defining Cloud Computing
- Part 3 – Know Your Cloud Consumer: ISV or Enterprise?
- Part 4 – Don’t confuse SaaS with Cloud Computing – Cloud Conference Week
- Part 5 – Cloud Standards and Open Source
- Also, I actually detailed most of my thoughts in last week’s IT Management Podcast. But for those of you who can’t put with that medium, there’s plenty of text in the blog here.
I originally had this post as one giant one but thought I’d divide it up, tell me which way you prefer and/or if this dividing up works better for you, dear readers.
The Conferences Themselves
As part one in the Cloud Conference series, though, the rest of this post are some notes on the conferences themselves. I’ve already written up my thoughts on SalesForce’s Tour de Force, down in Santa Clara, which makes a good starting book-end.
I missed the first day of O’Reilly’s Velocity conference while at Tour de Force. Based on the day I was there, and from what other told me the theory that’d evolved about Velocity turned out to be true – namely, that it was for the makers out there. Nothing wrong with that at all.
The slight mismatch in cloud expectations was that Velocity was largely about web performance. Indeed, as the sub-title suggests it was a “web performance and operations conference.” There was an interesting straddling between those two concepts – even sponsors split between what I’d call classically web operations and cloud computing concerns.
The sessions themselves were detailed and technical – from my stand-point, at least – esp. when it came to commentary on exactly what “web performance means.” Indeed, much of the discussion was definitional like that: here is a walk through of load balancing, here’s how you can do configuration management at scale, here’s war-stories of people having to scale up their performance.
There are some select videos up right now, and I hope they managed to record each session. For example, the panel on Surviving Scale was most interesting for anecdotes and stories. I always find these conversations most interesting for what’s not mentioned more so than what is mentioned: here, namely, ESM.
RedMonk client Hyperic announced CloudStauts here, and the two Puppet guys seemed to be getting a lot of attention. I also had the chance to talk with 3Tera about cloud standards – more on that later.
I must admit, I didn’t manage to go to any of the sessions at CloudCamp. Instead, I willingly got ensnared in HallwayCon with the finger food. Somehow, I managed to miss getting a beer.
From what I’ve heard, CloudCamp went extremely well: there were over 300 attendees (it was crowded). The sense I got from folks I talked was that it was equally as low-level as Velocity, but even more focused on actual “cloud computing” concepts and technologies.
I heard more than one person talk about how obsessed people were with defining what “cloud computing is.” Rather than get all despondent – as disucsions in RIA land along these lines can get – this is a good sign: once people start asking what a new technological concept means they (usually) want to believe in it rather than dismiss.
Overall, there weren’t a lot of clear answers or immediate results, but the attendance clearly shows organizers struck a chord with the emerging cloud computing sector.
I suspect I would have liked those talks quite I bit had it not been for the cheese ensnarement.
Also, see the sum-up from Reuven Cohen, one of the organizers.
The next morning I jumped over to Structure, in the weird, freshly built out area south of San Francisco (try catching a cab down there, buddy). As a perfect foil to Velocity and CloudCamp, Structure was the cloud conference for people who dress nice. You know: The Business.
Much of the conversation was at the non-technical level of discussing the “opportunities” of the cloud, broad questions of “how it changes things” and other concerns that above the level of “how we’re gonna get this stuff working.”
This conference was packed, so much so that I didn’t manage to really get into too many of the talks. There were press tables graciously provided (read: power hook-ups), but even those were packed by those faster than me.
After finding a hidden power-strip, I managed to sit down for some of the afternoon talks. They were a mix of boring, sponsor talks, Internet networking innovations, and then the VC panel.
As I mentioned on a the IT Management Podcast last week, it seemed like most of the VCs on the panel were no interested in cloud computing: either they’d already put their money in on previous investments, or had moved onto what looked like bigger bagger investments…or, they weren’t telling their secrets.
As happens with all confrontations between gold-holders and innovators, there was much gentle spatting between the panel and the audience.
Also of note was Greg Papadopoulos’ talk. Being a close Sun follower, I’ve seen his talk several times, so I thought it’d be the same one. But, much to my delight, his talk was newly updated with cloud talk, including some nice categorization. During the Q&A he got into some nice razing about there being no real standards, or open-anything, in cloud world (yet).
Hopes for Next Year
Missing from all of these conferences was the “we’ve been doing cloud computing and here’s how it’s been going” you’d love to see at trend-centric conferences. Really, I don’t think you could expect too much of that, this year though. If the cloud computing really does become something, by next year you’d want to see several of those sessions, along several lines:
- Providers – we’re providers of cloud computing and here’s what’s been working and not working. Hopefully Mosso, EngineYard, and others will be ready by that time to talk.
- New Tech – We started out new application on the cloud, building a SaaS, if you will, on cloud computing infrastructure.
- Legacy Tech – We moved our on-premis applications to the cloud.
- Customers & Users – We’re customers of people use cloud computing, or even SaaS, and here are the things we do different than when we were customers of on-premise.
- Raining on the Clouds – here’s when cloud computing makes no sense and, better, how it damaged companies and people.
On the enterprise angle, my chief concern is with the legacy angle: is it even possible to get the bulk of enterprise software moving in the cloud direction, or are they just stuck with, at best, consolidation via virtualization as they wait out the end-of-life?
On the bright-side, as I opened with, my sense is that we’ll see much of the next year’s (coded or announced, that is) green-field development deployed on the cloud. The economics and trendy appeal to startups seems to high. Hopefully, for the purpose of getting more information about people actually deploying on the cloud, any companies doing that will start talking publicly over the next 12 months. Otherwise, all this cloud stuff will still seem like a bunch of technology hunting down desire, which usually takes a lot of cash and time to pin down.
Disclaimer: Sun, Hyperic, and Reductive Labs are clients.