Catching up on some acquisitions and cloud announcements. After a little hiatus (I was on vacation for a few weeks), John and I are back, with pop-in guest Mark Hinkle.
Download the episode directly right here, subscribe to the feed in iTunes or other podcatcher to have episodes downloaded automatically, or just click play below to listen to it right here:
Show Notes
- CloudConnect – Mark and John were there. Meeting people, some good talks. Reprising the dev/ops, cloud/whatever tool-chain discussion. William’s SaaS chopping cut. Also: the “Farmville is killing our culture” article.
- Nimsoft, 3Tera bought by CA – $350M cash, some commentary from a NetQoS-cum-CA fella Patrick Ancipink. CA’s emphasis on mid-market focus for Nimsoft.
- Somehow, we end up talking about the success paths for cloud startups like Eucalyptus. We draw a distinction between “classic cloud” companies and newer, private cloud companies. The first focus on cloud as we know it, the second on strapping together virtualization, automation, and self-service to optimize and (we hope) make better IT Service Management.
- Dell and the cloud – Ubuntu & Joyent partnerships around PowerEdge-C hardware. Dell’s Mark Cathcart talks about the PowerEdge-C; he also collected together other coverage.
- IBM does dev/test as a cloud service – “lab management” along the cloud maturity curve as with virtualization. Cf. Austin’s own Surgient. John kicks the tires on IBM’s cloud here, and likes it.
- Speaking of CA, Andi Mann gets skeptical on dev/ops, and a reply from Andrew.
- OpsCamp Boston, April 22nd – if you like this podcast, you’re like this event. There’s one in San Francisco coming up too, May 15th.
- Several releases we didn’t get to talking about: AccelOps 1.6 release, Spiceworks 4.6, Versiera
Transcript
(As always, I don’t fully check out and correct transcript for this show, so if we’re saying something nutty below, be sure to check the audio before you think we’ve gone off our meds.)
Michael Cote: Well, hello everybody after a long hiatus due to my vacation. It’s the March 30, 2010, and this is the IT Management Club Podcast, episode # 69. And as always this is one of your co-host Michael Cote with RedMonk available at peopleoverprocess.com and we have the usual host co-host with us.
John Willis: John Willis at johnmwillis.com, but can be found at opscode.com.
Michael Cote: Yeah, I’ve noticed every since you’ve got a real job there hasn’t been a, the flood of information that there used to be on John M Willis, you must be actually working.
John Willis: Yeah, that’s right. I’ve actually I haven’t, I guess keep track of my time and actually work, I can’t just do whatever I want whenever I want anymore.
Michael Cote: Oh! Yes and we have pop-in guest who happened to be here in the neighborhood you want to, you’ve been given on a few times before, you want to introduce yourself?
Mark Hinkle: I’m Mark Hinkle form Zenoss, the community dude.
Michael Cote: Exciting, well you know, I think it’s been, let me see it’s been since the 1st of March since we recorded something. So there is a, you know, while I was in vacation I noticed some little exciting news items pass the twitter wire as it were and there has been some other stuff. So, you know, I don’t think we’ve got an overwhelming amount of things to go over, but there is some, this is an interesting mix of just straight up PR stuff and or I should say announcements, you know, just things that, events and then some, some commentary that we can add to things. But both of you’re at, I forget when it was, but you were at the Cloud Connect Conference, you want to, you want to like give us an overview of what that was like?
John Willis: Sure, we did. The Cloud Connect, it was pretty cool. I mean I think it was great because it’s they use that term [?], but it’s everybody who know in the cloud on twitter was there. So all these people that I never met, you know, I follow or they follow me on twitter. So it was really a blast for that. I mean it’s one place to get, you know, with so many faces with twitter names, it was really cool and the content was good, we did, we had a, that Monday we did a boot camp, CloudOps Boot Camp and so that was pretty cool. We had the guy from Zynga, the FarmVille app, the CloudOps security guy.
Michael Cote: Over the vacation I read that apparently FarmVille is killing our culture.
John Willis: Killing our culture.
Michael Cote: Yeah, you know, it’s kind of like video games are killing our culture updated to the FarmVille.
John Willis: Whoever wrote that article better get used to it, because I don’t see that changing anytime at least before it gets better.
Michael Cote: That’s right.
John Willis: I keep talking to everybody is like when I grow up I want to write a Facebook application, you know, so yeah, you know, he says, I said things are going pretty good over there, right, he’s like, “it’s pretty insane with your infrastructure and all that.” Because I’ve talked to some of the guys there, you know RightScale is doing some work with him, RightScale uses chef, but so, but the thing is funny they, he says, “yeah, John it’s very interesting,” you know, because I was the moderator for the day’s, it’s, you know, people give us real money, we give them back fake money and then they spend that fake money on us because it’s great.
Male Speaker: It’s like insane.
Male Speaker: But he did a great presentation on cloud and then we had a, James Urquhart, actually James Urquhart is like picking up the dev/ops banner, you know, he is a, he had a really cool presentation he gave, and I think he is written some articles I haven’t got a chance to read them yet but this whole, you know, the new dev/ops delivery, you know, moving from a service centric to an application centric model, and he is real–
Michael Cote: He is still, he is still a Cisco cloud guy right?
John Wills: Yeah, he is still Cisco, but he is really picking up on the whole, you know, what this dev/ops really means and how it starts this discussion of thinking more about the application and idea, it was really good, it was great, actually a great presentation.
Michael Cote: Yeah, well I mean from the stuff I read around it seems like there is, there is the big sort of like second wave of push of like Cisco Unified Computing sometime this year, probably in the summer or something and so, I mean obviously you would, that would be the part of the story they would go out with this, is you know Cisco providing cloud stuff, so that’s a good seating for all that.
John Willis: Yeah, then me and Mark did a little bit ass kicking on our panel right Mark?
Mark Hinkle: We did. It was good.
John Willis: It was a tools chain, we did a kind of actually it was Damon Edwards at DTO which, basic ControlTier guys and me, Mark and Damon, we submitted the proposal originally for the three of us and then they made us bring in some other guy that he had nothing to do what we were, we were trying to built that whole tool, tools pipeline or orchestration. We kind of had it really, you know, to a certain extent in sync on what we going to deliver and then out of the blue they had some guy with something like cloud applications, cloud r us and talked about how he can cloud everything with one cloud button.
Michael Cote: Yeah, that’s nice.
John Willis: So I thought yeah, that was, you know, if you were like doing drugs and really getting into the whole cloud thing you would love this presentation but for the real, but they kicked us, stood out like a real, a sore thigh, I know Mark, I thought we gelled pretty good right and the presentation was excellent.
Mark Hinkle: I think it was good I think it was just nice that, I mean there was probably like 70 people in the room and we are all talking about how our tools needed to work together and make it easier for them.
Michael Cote: Yeah, it sounds like it was a continuation of the OpsCamp thing, there was a meeting kind of along those lines. And so it seems like it was kind of that same vein of like we need to figure out the tool chain of doing this cloud inspired driven dev/ops whatever, we need to figure the evolving tool chain of the new way of doing IT service delivery.
John Willis: Yeah, actually Damon has been driving this hard and so like I actually got a call from one of the guys from the TechWeb to be there, so I said, gees wouldn’t this be great to just like take what we did at the OpsCamp in Austin and so really I’ve forgotten that actually it was kind of like connector, so you know, just keep moving along with the whole Ops.
Mark Hinkle: So yeah CloudConnect was good. It was a, it’s a great, you know, it was interesting for me now these days now I’m looking to talk to a lot of customers and I did get to meet a lot of customers there I know I think maybe, because there is so many [?] and there is so many on the business side kind of either abstract the customers or, you know, our customers seem to be a little hard to get to.
Michael Cote: Well, you know, you being the community dude Mark you come across the spectrum of conference attendees a lot and like, do you think there are a lot of customers or users if you will who are going to conferences at the moment or like what your sense for that like is it kind of, is this their economic problems, you run into vendors who have budgets for this more or like what’s going on there?
Mark Hinkle: You know, for us we try and hit the local shows, so like John and I were at Scale, which was Los Angeles and we’ll do Texas Linux Fest is coming up here in Austin and, you know, that’s, that’s the hypothesis. But when you get a Cloud Connect, I mean the people that were in that room I think for our presentation we did a little oh! There were, for the by and large I think there were cloud users, they weren’t vendors for the most part and that was surprising to me the number and actually the quality, I mean we were right in Santa Clara and so were we in Silicon Valley and people could drive take Caltrain or how are they going to get there and it was easy.
Michael Cote: Right, right now that make sense and so what was the part of the presentation you were giving Mark? What were you, what was the main segment you were looking up?
Mark Hinkle: Yeah, I was jut talking about how from the monitoring perspective, you know, it’s — we monitor this, we have all the data, but we got to make it actionable. And if you have service levels, especially service levels where you are looking at four or five 9s, you don’t have a lot of time to be alerted and make, make repairs to whatever infrastructure is down and still adhere to your service levels in a month’s time or year’s time. So the idea of automation where you have some data, that data becomes immediately actionable by a another tool whether it’s config management or an orchestration tool that knows to restart a service or things like that are really important. I mean right now the solution is to run to develop very, very redundant systems, the fall over probably as a back up to that, then you need to be able to make sure that your monitoring system and the data was actionable and that we could kick off processes in another tools.
Michael Cote: Right, right I mean it sort of the basics of just kind of knowing, knowing what’s working and what’s failing, so that you can actually do some management base stuff of what’s failing obviously, I mean I usually when something is working, I guess you don’t really have to touch in very much but–
Mark Hinkle: So I got a question for both you guys how can you guys say ‘data,’ isn’t that ‘data.”
Michael Cote: ‘Data.’
Mark Hinkle: ‘Data.’
Michael Cote: I thought I said ‘data’ do I say ‘data?’
Mark Hinkle: I thought you say ‘data’ sometimes.
John Willis: To me it’s that — it’s that Pennsylvania Dutch sort of messes me up I say ‘crick’ instead of ‘creek’ and or it word wash.
Michael Cote: Or like pecan versus pikan or pikaan, pecon.
Male Speaker: PK?
Male Speaker: There you go, I don’t know, I don’t even know the three ways of saying it.
Male Speaker: My dad used to say ‘earl.’
Male Speaker: For URL?
Male Speaker: No, no I mean I don’t think my dad ever saw a URL; he died long time before URL’s came into the world.
Male Speaker: He had to deal with IP addresses.
Male Speaker: Yeah, yeah, right there, no but he used to call ‘oil’ ‘earl.’
Male Speaker: Yes, of course.
Male Speaker: Who is to say also?
Male Speaker: No I mean I like to think that I say data, but I guess that, isn’t there an English way of saying it too that’s like ‘daa-ta’, there you go it’s data versus ‘daa-ta,’ that’s a slight difference. I really have no idea.
Male Speaker: What did they know, the English guys you know call soccer football.
Male Speaker: Yeah that’s crazy.
Male Speaker: What does foot and a ball have to do with that?
Male Speaker: Yeah that’s right.
Male Speaker: I went South Africa in the fall and I now say ‘Zaabra.’
Male Speaker: Oh!
Male Speaker: Yeah, I was told the appropriate pronunciation for the animal and since they have zebras and the U.S. doesn’t, I guess they’ll be able to get their pronunciation.
Male Speaker: We’ve only got zebras in captivity apparently.
Male Speaker: So you know the thing that I saw from Cloud Connect that seemed interesting was our friend, I phoned William over at Oracle. He was trying to cut the SaaS off of the cloud stack with a guillotine, which I thought was fantastic little slide there. It looked he had kind of a micro key note that was just, he had a great slide. I have to put a link to this, it was basically like here is all the, you know, the 20 current models of how to talk about cloud stuff and. And then the last slide was a, like I said guillotine cutting SaaS of the top of the stack of cloud stuff. His point being if I, and you can since you guys were probably there, you know may be saw and said to me just reading his own post of it, his point being that, it’s, there is not much distinguishment between a web application and a SaaS application. So it’s kind of silly to make SaaS a special sort of thing instead of just talking about web apps and the more interesting stuff from his perspective are the other two layers and the other one is just sort of a consequence of those I guess.
Male Speaker: Yeah.
Male Speaker: What, what was the last thing he was going down?
Male Speaker: I didn’t see the presentation, I didn’t see his presentation. I was kind of, you know with the two presentations, one day I had to give and other presentation, but I did meet him. But I don’t know, I didn’t go back and take a look at it, I don’t know if I agree with that, you know, there is clearly a, I know, you know whether SaaS is cloud or not, I don’t give a rip right, but to treat it differently or define it differently as a web app, I think that, I think it should be. I mean I think that there is a difference between a pure service that runs, you know, kind of adds the service from just a generic web app, you know what I mean? I think I am not sure that I make that connection but I do need to go back.
Male Speaker: I mean, I guess you could say, you could have, you could a web app that doesn’t, doesn’t take advantage of cloud technologies right, like the, you know you could have sort of like a default or an older WordPress or Drupal or whatever installed and it doesn’t really you know do anything cloudy. It just sort of, it’s isolate on its own essentially and just so that there could be that sort of argument. But I mean that implies that there is something about, something that calls itself software as a service that means that it’s you know sort of scalable and all this other coupon automatable and so forth and so on.
Male Speaker: Yeah I think it’s clearly got to, I mean again I think the whole, like we’ve gone over this a billion times but what is a cloud and what is not a cloud is just beyond silly these days right?
Male Speaker: Right.
Male Speaker: Because the, just you know, but I think that the, the model is important. You know whether you call it a cloud or not, you know that there is a model for and has a service or Google apps or you know or you know salesforce.com, things that clearly you know are different than just this plain web app, now web app can clearly be assessed but and then you get into the platform as a service, an infrastructure service and those are you know cloud but I think that, I don’t know, I don’t know if I’d agree that would chop SaaS off.
Male Speaker: Yes, yes, you know reading the, reading some of his own commentary and stuff around it, I think what’s, it’s, the interesting thing that it motivates is something that I’ve tried to do more nowadays in the area of cloud. And not really, I mean it may sound kind of absurd but I try not to focus so much on what it is but like what opportunities it enables and like what sort of things it enables someone to actual do, because they are using these technologies, because you know kind of like you are getting, and the truth of the matter is this, there is a, in same way that there used to be lots of different weird operating systems and spreadsheet programs and things like that. There is a whole bunch of different things that are not exactly the same, that kind of fit in the space essentially. But they are all essentially trying to enable the similar bucket of things and I think that starts to become the more interesting conversation, which you know it kind of, there is a bit of fragmentation and what not going on rather than you know unified standards and blah-blah-blah. But at this point I think this cloud crap has been going on long enough that it’s time for you all just to figure it how to use it and not really, you know, struggle with what it is.
Male Speaker: I love, Damian Edwin since you know, think of like would you know think about like, like I try to describe things and he is the one who got me thinking, he’s like you know, it’s like saying you know, Computerworld, the magazine Computerworld today you know, a lot of times when people say things like you know, “I don’t know if it’s a cloud and only a cloud if it does this,” and I say you know change the word ‘cloud’ to computer and or Internet and think about how silly that sounds today. It going to sound just as silly in about three or four years from now, you know what I mean?
Male Speaker: Yeah we’ll have computer computing, I’m looking forward to it.
Male Speaker: Or, I just you know that, like right now a magazine called Cloud World would be unbelievable, only with their right mind would start a magazine that it didn’t exist already called Computerworld.
Male Speaker: Yeah, yeah, yeah.
Male Speaker: Internet World, oh, everything that’s going, what you needed to know about the Internet, you know, up to date info about the Internet, Internet World, you know, get your subscription now, I’m, you know that’s, the cloud is like you know, it’s almost a [?].
Male Speaker: So on the track of, this sort of like one foot in this, the cloud and the other, the other foot in old stuff, also over the little break that I was on, CA bought another company. So in addition to having 3tera they purchased Nimsoft for, for like, according to their press release for $350 million in cash. There wasn’t any like funny paper or anything trading, it was just dollars which is exciting for Nimsoft I guess. And it seems like the, I mean to give the quick summary before we get into it, the line of the motivation was one, two, there is sort of three things it seems like. One of them kind of speculation more or less, but the first one is like, you know, getting in more into the mid-market and emerging new uses of IT monitoring and management stuff is what the press releases from both side said. And then there was also Nimsoft has been talking about how they have had good entry into the managed service provider MSP space, which is sort of like an old school pseudo cloud something or another. And then also the other thing I think what with the near acquisition to 3tera and maybe some like [?] here and there, which maybe opportunistic or maybe real. There is some notion that, you know this sort of, that the Nimsoft stuff will help them with sort of cloud monitoring and management, the same way that. I think people were looking towards Hyperic to help SpringSource and VMware out with that kind of stuff as well to make a call back to the past summer here. But so what have you guys, I mean I’m sure you guys looked at that what, what are your thoughts of it?
John Willis: I think it’s a game change, I think it’s going to put CA on the map; it’s going to change everything as we know it today. The integration of 3tera, Nimsoft and 1000 other products they pulled over the last 30 or 40 years, they are finally going to get them all put together and it’s just going to kick butt. In fact I think it’s going to be a SpringSource and CA just up there and neck and neck at the finish line.
Male Speaker: Oh John. You know there is like you know for as, along those lines I always, I always recall CA having, you know CA is notorious as John is alluding to with his tongue twisting comments for having a giant portfolio like a massive one and you know I remember back when I, it was about four years, when I just started at RedMonk I remember there was a big announcement from CA that they were consolidating their brands down. And they released a PDF that had the new brands that were listed there, and it had gone from like a 100 page PDF to like a 60 page PDF. And it was, you know, kind of a funny illustration of the point, but yeah. It’s, the CA is known for having a lot of things and you know they had actually bought Ultra Point an Austin based company last year as well. They actually purchased a lot of things last year.
Male Speaker: Yeah, well they you know if you think about it it’s like cosset right, which was the cloud computing and then also now they got 3tera, so what do they do at cosset I mean, you know, I mean arguably I don’t know it’d be interesting. I mean 3tera has as a lot of potential. I always felt they had a lot of potential, you know somebody said that, you know a rumor at Cloud Connect is they paid a 100 million for 3tera which is you know, where 3tera should have been a $100 million, it is like you a bail out, you know what I mean?
Male Speaker: Right, right.
Male Speaker: Because they should have owned this whole, you know, they are still they are kind of modelling canvas for doing, building cloud you know multitiered infrastructures, because nobody else has it you know, but they just really didn’t take it to where they should have. And I don’t know it will be interesting to see, you know, CA is not great for redesigning or redeveloping products. And I think that from what I know about 3tera it’s got to be a lot of, there is going to, have to be some real massive architectural development resources put into that to make it a game changer and I’m not sure, now that it’s part of a big company. Do you think, a little company has a better chance of doing that than a big company in my opinion, right?
Male Speaker: Yeah. So do you think then the opportunity is there for Eucalyptus?
Male Speaker: It’s a good question.
Male Speaker: I mean just the fact that, and I agree with you I think that you got to win some of these battles, you got to become pervasive I think it’s hard when it’s part of –
Male Speaker: I think this is like VMOps, these guys, the VMOps’ the new guys that are coming out of that, you know, they‘re doing it, they’re, they’ve learned from all of the mistakes and they’re, you know, there’s a whole breed now of these new, there’s a window for, you know, Eucalyptus you couldn’t have given it a better try right. So Eucalyptus had the money, had the momentum everything I’m just not sure that they have met the expectations of where, you know, what I would have thought they should be right now and, you know, or I would say certainly with Benchmark Capital Investment, I can’t believe the Benchmark Capital is happy as they’re with Eucalyptus as they’re out of scale. Just being [?] and so I think, I don’t know, I don’t know I can’t tell you whether in my opinion whether Eucalyptus dropped the ball yet or not, you know what I mean? I hope not I mean I like those guys a lot and I think they got a good product, but I’m seeing these guys like the VMOps and there’s three or four new players on the market that didn’t exist this markket, not Marketta they’ve changed names like 99 times and it’s another —
Male Speaker: Yeah, that’s a long lines of what their new name is.
Male Speaker: Yeah, yeah and, you know, what I’m talking about right?
Male Speaker: Sure.
Male Speaker: It’s – yeah so –
Male Speaker: The OSS 1700 or whatever.
Male Speaker: Yeah, they it started out as OSS 1700 and then who’d they become then they, it’s or they’re not going to become OPS guys right?
Male Speaker: No, no — yeah, yeah those are the same guys.
Male Speaker: They’re the same guys.
Male Speaker: They were in that, in VMOps now they’re like some Sanskrit name or something.
Male Speaker: Okay, yeah so and there’s a couple of other companies like that that are emerging that are just getting a lot of traction and showing up all over the place. So, you know, like I didn’t – like Eucalyptus wasn’t at Cloud Connect. I don’t know whether that means anything or why weren’t they but, you know, but there were some other lot of cloud players that are out there. So it’s a, it’s a bloody area to be in the private clouds space right now.
Male Speaker: Yeah, yeah it’s seems like there’s this, there’s this at the moment unresolved tension between establishing like virtualized self-service run-book automated infrastructure right on the one hand as a private cloud. And then saying like we really want to just replicate what happens in the public cloud or in private cloud. And I think you need that second one to win where people are going to be like, “all right we’re going to build, we’re going to build a cloud that’s not just, that isn’t just a bunch of, and not to demean it or anything at all but, you know, that isn’t a bunch of automation and virtualization and in the IT management policy stuff wrapped around it right.
Male Speaker: Yeah.
Male Speaker: You know, they’re sort of like those are the two things I tend to see at the moment is what IT service management once it’s evolving to be a little to be more efficient hopefully or there is like this big sea change that you’re going to run to clouds right. And I don’t, I haven’t heard a lot of people who are, I haven’t heard enough people I should say who are like, you know, they’re just kind of sort of like do a scorcher thing where they just install a cloud or something. And so that if you’re a someone who like Eucalyptus or whoever who basically has a cloud for a framework then you either looking for a giant buys from MSPs and people like that who want to run a public cloud or enterprises who want to do cloud stuff as well. I mean, I think it’s still pretty open at the moment like what that’s going to shake out to be but it’s a, it doesn’t seem like the market demand has necessarily happened as quickly as everyone would like it to.
Male Speaker: And so you always say things much better than I do but I agree a 100%. And I think to me that the thing about Eucalyptus is they were banking on what you just said. They were banking on this kind of disruption of people, you know, and I think they put too much banking, “Boy if we look just like Amazon it’s just going to be awesome for us because anybody who is already getting beat up on Amazon from an enterprise perspective, it will be a no-brainer for them to go to Eucalyptus for, to be able to have that kind of hybrid, this capability of a hybrid solution.” And the truth of the matter is what I find people are finding is building private clouds is really not, you know, your basic private clouds these days really pull down to self-service provisioning. You know, people want the ability to give their developers or their testers or QA or HR department or something the ability to get resources without having to interact with human and there are lot of, to figure out how to put that glue around some virtualization solution for a OpenVZ or it’s, you know, Xen or it’s Xen Citrix or it’s the VMware and they’re perfectly happy, you know, “you’re going as well you can’t run it that way, it doesn’t have this robust type, you know,” “What do you mean? Our developers are happy, I’m happy, you know, we’re using a stand, we’re using the virtualization platform that we already have a lot of IP in, you know.” I don’t, I think that, I don’t think that the Eucalyptus is as the world expected, you know, I or maybe they don’t know it but that’s happened really I mean I went to Pulse they were like six companies at IBM Pulse that were talking about their private cloud and all it was a hybrid of [?] VMware.
Male Speaker: Yeah.
Male Speaker: And, you know, they were calling it to cloud, they were happy and Eucalyptus ain’t going sell them a dime more for software.
Male Speaker: Yeah, yeah and, you know, I mean that gets to like when, as boring as it is like when I talk with people who are clients or not or whatever right and they have some cloud thing, I’m always trying to encourage them they sort of pay attention to the enterprise market and enterprise needs here. And unfortunately that means a really slow transition that builds on a lot of what exist already right?
Male Speaker: Exactly.
Male Speaker: I think that’s why you see a lot of success of like I mean we’re speaking kind of like diminutive, yeah but actually is an improvement of things to have, you know, the self service automated virtualized, whatever private cloud stuff right and all of those concepts existed before and were well explained and understood and maybe the technologies were good or maybe they weren’t. And so it’s easy to say like well all of these things come together to be with this private cloud is that you’re interested in, and so you kind of advance that way. And, you know, eventually hopefully you get to like what’s more of a pure cloud sort of idea. But you really need that unless you’re going for that other sale where there’s someone building a cloud from the ground up, right, which is a whole completely different market that might have its own sort of stuff dynamics or does I should say. You know, you need to analogously build in the enterprise data centers from what already exists because otherwise you just keep people who, not, I don’t know it’s just doesn’t, it’s sort of like you got to jump, you got a Wile E. Coyote it over a canyon and, you know, realizing your falling soon enough I guess.
Male Speaker: Well, it is a good lesson why I never underestimate IBM I know that but some people do underestimate them and, you know, this when there’s something that they want to own they’re not going to lose it, right? And so, you know, a year and a half ago the discussion on private cloud, Eucalyptus looked like my god this thing could win at all, IBM basically put a stake in the ground two years ago, so we have a cloud right? So that was an immediate classic IBM so like say they have something before they have it right. But what they’re doing is, they’re setting the boundaries of their ownership right and now you look it again, you look at their cherry customers BofA and JPMorgan Chase, the large banks and all those guys they’re running what they call a cloud. You can laugh at their clouds but the thing is they’re not laughing at their cloud and, you know, in IBM, you know, made it so, you know, and I don’t know the HP world as much or, you know, but –
Male Speaker: But you know, these big companies, you know, like you just said they’re going to stick with their, you know, it’s easier to convince people to say, “hey I can get your cloud this way or go out and buy a whole bunch stuff you’d never seen before.” So, you know, and so and really—
Male Speaker: Yeah, and that’s the other problem with the sort of classic idea of what cloud computing is as odd as that sounds faces and the marketing –
Male Speaker: That’s right.
Male Speaker: Is everyone, everyone old and new has a cloud offering at this point right. And so you could be waiting for like vSphere or you could be wanting vSphere or you might want to know what Microsoft is up to, or IBM, or HP or the new companies right? You could look towards RightScale combined together with Amazon or [?] I mean there’s a whole lot of choices out there right. And so at that point it becomes a marketing challenge to sort of get people to want your technology, you know, apart from the technology itself, looks like it sounds terrible. But when you have like that kind of saturation out there there’s it’s just it gets to be very difficult no matter what your underlying technology is. And, you know too, in that area if you’re one of these cloud start-ups a lot of what you’re doing is trying to establish yourself to get bought by one of these companies. Because eventually when it does happen with large companies for the most part and, you know, it’s not a blanket sort of thing but they tend to, you know, acquire by innovating. And so what they look to do is kind of slowdown the brakes, you know, slowdown the market enough, so they can kind of like capitalize on it themselves and sort of figure out what to do essentially. And I think there’s a certain Stockholm syndrome where some people in the enterprise like this and some people don’t like. It just depends in which way you hang I guess, so to speak and, you know, so what happens with a lot of, probably what will happen with a lot of cloud start-ups is when large companies who became – when companies who became successful with their vision of whatever cloud computing was, whether they’re large or small is, eventually they’ll need to deliver on the rest of the roadmap where they want to, and they’ll purchase of these smaller companies and these smaller companies will start to finally, they’ll be at the right part in the timeline to deliver what it is they’re delivering. And, you know, I mean I think 3tera is an interesting example of this because they were around for a long, long time relative to everything we’re talking about here. And, you know, I think CA were then speculating rather, but I think CA was like, “oh, we need some cloud stuff,” and it sort of rightly sold right? And they’ve probably, they’ve probably looked around at everyone, as everyone who is acquiring would do, and I’m sure through many different combinations of both technology working and a good price and an existing customer base, brand value and all this, that they looked at, you know, 3tera was their choice for someone to instantly get them in the cloud game, if you will. So that’s probably what will happen with companies at the moment, who are successful or not successful.
Male Speaker: Yeah, and 3tera actually had a pretty good penetration in the kind of hosting provider space. So, you know, if it’s true that it was around a million dollars to just get, you know, to resell other products to those customers that hundred million dollars could be a throw away, you know, what I mean? I’m not saying it is but–
Male Speaker: Yeah, if that’s true that was a good deal.
Male Speaker: But Mark I got a question for you on, so speaking of g-Eclipse I guess that Marten Mickos now, so if I’m right that they have dropped the ball, and I think I like once you get in all those guys, I think they’re awesome but they just announced the guy who is the founder of MySQL who I don’t really know much about but they’ve made him the CEO, so maybe he can make things happen, I don’t know.
Mark Hinkle: Yeah, I think he is I mean he did a great job with MySQL, what’s interesting is, they took MySQL to Sun over the course of 13 years and had the luxury of building, you know, user base and a brand to convert and as fast as cloud computing is moving, they may not have that luxury of time, but I really think he’s a pretty capable guy and well respected and brings a lot to the table there.
Male Speaker: Yeah, you know, I mean he — Marten Mickos is, he’s part of a handful of like, I don’t know what you’d call the modern day open source start-up CEOs that are very well respected and obviously do a good job. I mean there’s other people, you know, like Mike Olson over at Cloudera and people who have been through, been through the open source ringer of acquisition and starting up other companies and it got generally well, in the case of these two, they were always very nice guys and yeah, I mean I think it’s hard to say to anything about Marten Mickos, you know, because he had — he has the understanding obviously of open source dynamics and also – but he also has that very pragmatic, “we’re here to make money” sort of thing without being sleazy about it. That can put a lot of people in the open source world off, so, you know, it’s a – we’ll see what happens.
Male Speaker: I think it’d be interesting—he brings some of the old gang back together from MySQL, seems like there’s a lot of refugees from Oracle these days and like Zack Urlocker, sorry my Pennsylvania Dutch again. He had a really interesting post on his blog about like the marketing benefits of open source for Sun and some of the things that they accomplished last year but, you know, that kind of horsepower with Eucalyptus, they could, you know, get out ahead of the race pretty quickly.
Male Speaker: Yeah and, you know, I should since we’re talking about this so much, I should disclose that Eucalyptus is actually a RedMonk client but that’s, that is one of the strengths that have is — and then also building on the MySQL thing and, you know, I mean Marten knows all about this kind of marketing but they have a good product that people can self market to themselves in trial and try — out and everything that, you know, theoretically builds a good funnel to paid things, or at least growing a user base which if you’re just looking to be acquired, you know, growing a community could be as good as growing revenue or it can be somewhat as good as growing revenue I should say. And so, you know, if they were sort of tried-and-true for any sort of open source based cloud thing or any cloud start-up, if there’s, if you can kind of grow the community of users that you have and it’s just sort of like a de facto standard to what people are using but it’s, you become an extremely valuable company either for being on your own charging revenue or for like I was saying being the target of innovation by acquisition thing. And that’s definitely like the story that MySQL made its final sale on and many years of success, which is you can download it and use it and then it sort of like leads into other value added things. And I think, I think the challenge, you know, generalizing the rambling on here, generalizing the open source I mean I think the things that’s always been interesting to figure out is how dynamics of operations open source is from development open source. And I think that’s something that, you know, with Zenoss being around for a while, it’s not quite as ironed out as like it was from MySQL and Middleware and stuff, like that model was pretty well understood whereas in operations I don’t know things are slightly different and I think that’s kind of the thing to figure out at that level.
Male Speaker: So now that we have been in that [?] of course.
Male Speaker: Well, you know, just before — they were going to call us the Eucalyptus episode, but I was just thinking, I mean their relationship with UEC I hadn’t really, now I’m circling all the way back they haven’t got much about it but economically you will see I think the fact that a purest open source guy who really understands the open source community is going to work really well with economicals, you know, integration with that product.
Male Speaker: Right, right.
Male Speaker: So it might be a win across the board for all of them.
Male Speaker: Yeah, you know, that’s a good a segue into like the little spattering of Dell sort of news, you know, the –
Male Speaker: Yeah, there you go.
Male Speaker: They had Dell sort of, I don’t know, if finding is right word but they had another sort of cloud announcement, if you will, that was, you know, it sort of, you can see that Dell sort of like and their partners are building up a cloud stack, if you will, so they have this they call it the PowerEdge C server, which I guess, it’s sort of their cloud oriented server and I’ll put in a nice summary link from Mark Cathcart, you’ve probably heard reference, heard on some podcast here and there. You know, he’s a former IBM guy that went over there to work on, I don’t know if it was this box, but I think it was, to essentially do the systems’ design for this and then on top of that, you have partnerships with Ubuntu, which like John was saying includes Eucalyptus and then there’s a, I think there is also a partnership with Joyent’s, Joyent sort of cloud management staff or something like that. So, you know, it’s always fun to see, for Dell like these kind of opportunities are all, it’s sort of all new territory for them as far as like building the stack up, you know, new things for them to get into beyond the hardware, so it’s always interesting to see what they’re up to.
Male Speaker: Yeah, I saw that I’ve put the announcement because Simon gave me a shot out on Twitter like, hey thanks for your work, so it’s good to see that. I did, I can’t talk about it, but I did a lot of work to help that deal and it was great to see that they actually, it went through so.
Male Speaker: Yeah, and you know, it’s I think and, you know, being completely you know off the blue sky in theoretical again right like this is, without, the role of open source here is like the same role that kind of Linux does for IBM right? Like Linux knows the somehow they know, somewhere they know the value that Linux provides them as far as overall sales for their servers and everything else and the same can be true, it can be set of Oracle and other people who are selling things right. And while Linux may not always be a huge directly revenues of monetizable thing, you know, you can buy a license for it and or you can it get it for free but it may not be a huge chunk of thing because it’s definitely a part of the stack that is in an enabler. And you could see with all the different cloud stacks or to put it in another way, next generation is data center stacks that people are trying to do. These various chunks of cloud stuff work could kind of fit in there is the thing that makes it possible. And then you’re in kind of the weird position of indirect revenue and all sorts of stuff like that but at least you can see that it’s sort of, these things start fitting together as far as a new way of managing data centers for you. Like I said, to get all excited about it.
Male Speaker: That’s right. So in other news I heard you missed it out that the IBM’s dev test stuff.
Male Speaker: So yeah the IBM public clouding, I guess which is the first real offering in the public cloud, so a cloud connect a, they’ve kind of made the announcement I think that week and now it’s all set to be to go cynical on, you know, because I’ve been, actually I’ve been a calmer, what you could look these days right, knowing that I work for a vendor, I can’t go about telling everybody is right, particularly when Mark will introduce me to the people that are out there in Australia right, Mark.
Male Speaker: Do you wanna tell everybody you have a new girlfriend?
Male Speaker: Oh I have a new girlfriend yeah Tara Spalding over at GroundWork’s, we are best friends.
Male Speaker: Oh, that’s good. She is quite lovely; she is a very nice person.
Male Speaker: She is hot, she is awesome; it was weird because I’ve been so tough on GroundWork-
Male Speaker: Yeah.
Male Speaker: And Mark is like you’re kidding me, at the Ops Camp, he’s like you’re a rotten friend that I will push into a China cabinet, you know, he’s like, hello Tara, here’s John, you know, like oh gees, but no, she was awesome, he’s got a picture of us hugging but, so I’m trying to be there the good guy but for IBM, you know, I mean just, you know, they’re IBM, and so everything is all right and I’m going to go take a lot at this cloud and it was adorned the presentation one the guy is talking about, the IBM cloud, you know, it’s like okay, you know, so I asked him a couple of questions and he even tweeted like, here they’re going to make a mass mistake and I wanted and they had all the basic ingredients you need for a puppet. I mean, image management, they got instance management, they got storage management, they got IT management and key management and that’s basically Amazon, one-on-one, right and the ability and it’s all API residual API, gets and posts and deletes and basically, you know, you want to create an image you do a put, you want start an instance, you do a get, it is just not a get but a post, but the thing is it has full API set completely documented, well documented and you can do all the basics of creating images, they don’t have Window support yet. So they’re like, they’re like where Amazon was two years ago or 2 ½ years ago, you know, but they did it right, I mean, it’s seems completely programmable.
Michael Cote: And just to summarize, it’s basically a virtual lab offering that they have that’s targeted at development and testing, that sort of that phase of release management of IT if you will. So it’s not necessarily, it’s not necessarily for running in production, but it’s for using as just a lab.
John Willis: Yeah, but then, you know, it’s a good that’s really what Amazon was for the first year.
Michael Cote: Yeah, and that’s how virtualization like got a lot of its initial footing and all sorts, I mean that sort of the sort of grounds of how you, you spread in the data center essentially.
John Willis: So why did, you could say earlier that there was not the legacy cloud definition or the classic cloud definition.
Michael Cote: That’s right.
John Willis: In your classic cloud definition, yes you would call a virtualization provisioning, but since API based and it could be completely driven without a human, I’m going to call it a cloud.
Michael Cote: Yeah, and now, I mean that the Sun had what did they have, one of their, I forget the name of the project, but one of the interesting aspects they have had in one of their development, their cloud host development platforms, it is kind of like their source forge. They were getting towards being a dev/test virtual lab and doing builds and things like that and there is a lot of interesting advantages to take advantage of especially in the development and obviously in Q&A and most of it being their like all of their different stupid servers, they’re not stupid, but it’s frustrating. All of the frustrating servers and configurations you need to test things on can be a huge bottleneck for doing things and so just having assistance with that is pretty fantastic, I mean a lot of the elasticity needs and spikes and all that kind of stuff that their early promises of cloud computing sort of came from, like being all the service, all the sales on Thanksgiving and Christmas and whatever other days are big. Like essentially fit the model of doing development in QA where there is a lot of busty activity and maintaining everything around it can be really freaking annoying and it’s also a very low risk way to get know cloud stuff.
John Willis: Yeah, I think it fits their model, I mean I didn’t go too deep into it, but they have got like images, they’ve already got a lot of prebuilt images around the WebSphere stack and the development stack for their, particularly and so they’re attacking this cloud like IBM attack everything else, slowly but done with the right framework. So, again I would have really criticized the cloud if you couldn’t do it, completely self-serviced if you couldn’t build around it. They did all the right things for a base one-on-one public cloud and catering to their audience, it’s going to be huge, I think it’s – I mean I could see a lot of possibility, you think about the whole business partner, ecosystem around IBM and all their products and what’s happened with Amazon it could be insane amount of money that people could be making just building their confidence around the IBM public cloud. I don’t know the thought is mind boggling.
Michael Cote: So there is a feel, there is like three other things, I had on the agenda to mention and two of them are related. So, you were I think while I was on vacation there was this little twitter thing between about dev/ops to speak of one foot and one thing and a foot and the other.
John Willis: What’s this dev/ops thing they talk about?
Michael Cote: That’s right and I think it was started out by a post by Andy Mann who was a he used to be in Enterprise Management Associates and now relevant to a lot of what we have been talking about works at CA, I think on their cloud stuff essentially or something along those lines, I forget exactly what he is doing, but I’ve got to know him over the years, he seems like a interesting guy and he definitely – from the posts he has been doing recently, he definitely does have nice interesting things to say about stuff. So, basically he had this whole, I’ll give my summarization of the post and I think you guys might have actually witnessed like the Twitter thing or whatever between him and Andrew Schaefer and some other folks, but he was basically saying, you know, he was very similar to the argument you would see about Agile way back, so you got this dev/ops things and it’s kind of leaving ops people holding the bag for all these crazy cowboy people and sprawl and process and security and locking down all these sorts of things and so I’ll put a link to his post, which I think was actually a very good like it’s a very good summary of all the concerns that aren’t going to go away that we need to make sure to take care of right and they are kind of like the boring and the weed stuff that you don’t see in the frothiness of all this conversation so much. And then there was a reply post from Andrew Schaefer who has been on here a few times before who is basically saying in classic reply style, the dev/ops does all of these things to the hilt. So, it seemed like a couple of people have mentioned about it, like you missed a good little Twitter discussion over the break there so. What did you guys see about it?
John Willis: Lesson number one, don’t bring a knife to gun fight. If you’re going to have a debate with a little idea Mr. Andrew Schaefer you better be gunned and gunned, that guy is a tough cat to debate on Tweeter. He has had me sometimes, so I just want to fly to Utah and just beat him up, but I don’t think I can, he is a tough, I mean Schaefer smart as a whip and, but so that’s number one, but two any link like totally it was like the cloud they say are one-on-one, right.
So, when the cloud came out, the first thing that people wanted to say is right like three page articles while the cloud can never work. It will never work because, here is all the things, the thing is dev/ops is a concept, right. It’s not a product, it’s not a tool and it’s not something that you turn on or turn off.
So his article was foolish from the perspective that what he was arguing about is like well dev/ops is it going to do this, this, this Andrew was spot on, like, I mean what you needed to — he was like peeking in the curtain and looking at what he saw and he only saw one perspective of it. He hadn’t been to a dev/ops stage. He hadn’t understood what everybody is trying to accomplish, spend a couple of days with some dev/ops guys and Andrew listed a bunch of people and talk to them about the problem they are trying to solve, go the velocity, spend the week with velocity and listen to what Flickr and Twitter and these guys are doing and how they are redefining how operation work right, and then step into the argument, because some of his points are absolutely right.
The idea of Agile infrastructure, Agile operations may not work, right I talked to a guy with a large 4,000 service shops and we talked about Agile operations. And he said, “I don’t know” he says, you know, and he is like “I don’t like this dev/ops thing.” I’m like, “it was the same thing all over.” He was like he had heard about it and he thought it was like a massive change in a tool or, you know, what are they trying to do, they don’t understand what I do, I’m a real busy guy. I do a lot of work. I managed 4,000 servers a year, we’ve got continuous deployment. We do a lot of cool things, and what I don’t want is a bunch of outsiders trying to, and we talked about and I said dev/ops is a concept, it’s a stake in a ground to try to address a problem that the industry has had with developers and operations not working as a single business, you know, or working to solve a business solution.
So if the word sucks who cares, it’s like cloud if the word ‘cloud’ sucks, I think the word ‘cloud’ sucks but if it gets everybody to start building clouds, it’s great and yeah. And so I think that’s what dev/ops, and when we had the discussion, the agreement was, yes you’re right John dev/ops is a good concept and it is a good thing but, because it is raising awareness, but at the end he said to me, he’s a guy from the shop, so he said, “at the end though,” he’s, “I’m still not sure if the concept of Agile operations or Agile infrastructure is going to work,” and then he started to explain why, and what we both agreed is maybe the concept of operations and how it works with Agile needs to be refactored, whether we call that Agile operations or Agile we don’t know.
Male Speaker: Yeah, I mean I spent a lot of time with some, a lot of time. I spent some pleasurable time with Agile development thinking people kind of like noodling on this, and you’re right that there’s, the touch points there aren’t extremely clear at the moment right. And I think the main reason, they are not clear. This is the way it was with the early Agile or early Agile. So it’s basically, there was basically the extreme programming book and that was it and there wasn’t like tools or anything.
So everyone had to kind of experience firsthand what this thing actually looked like when you’re doing it. And over the — I don’t know over the, I don’t know has it been 10, over 10 years I think since that book was out there, there’s plenty more out and there’s experience and, you know, this is just like the boring part of a technology concept where people have to just come out and say what they did and document and everything. And then the next stage is you have a bunch of tools that actually embody what all this process is. And so now in the Agile world with all these tools that, you know, they’ve got stories and all those stuffs that are parts of the way of, thinking about Agile.
So it’s really easy you can like get a tool that is Agile development as absurd as that statement is and kind of get a sense of what it is and what it looks like. And I think for us it’s exciting and as much as we go on about dev/ops there’s, it’s still difficult to say to point at numerous mainstream examples of like, oh this is this is dev/ops I mean that’s why, that’s why we always use Flickr as an example because that’s one of the only well documented or well enough documented to talk about once out there. And it is a, yeah, you know, you got to sell people on the potential of it being good stuff but it is, but it is sort of, yeah.
Male Speaker: But I think like you’re getting out when you kind of sit down and you — it gets back to that discussion we were having earlier you kind of have to admit that “Yeah, yeah, we’re trying to do the same thing we’ve always been trying to do. It’s not like we’re trying to do something new.
Male Speaker: Right.
Male Speaker: It’s just like might be this time and there’s different sort of things operationally you would do it to take advantage of this technology in this concept.
Male Speaker: Right, yeah. There’s a lot of different thing, you know, it’s like the same thing. The worst thing I ever hear, I heard here when I get some old timer and says, we started talking about the cloud, and “I saw, yeah cloud absolutely since we’ve been doing that for 15 years.”
Yeah. No, you haven’t but, you know, so if you wanted to say that you’ve been able to do, you know, self-service or virtualization or grid and yeah a lot of those components but this cloud had happened two years ago is different than anything, the instantiation of what the cloud has produced or, is significant and it’s changed. And I think the dev/ops is at that, where cloud was three years ago, dev/ops is at that kind of place and people even looking at it as, “oh I get it.”
I think if you don’t get it you have never lived it right? Anybody who has had to deal either from an operations dealing with dev or dev dealing with operations, they get it immediately. It’s like oh yeah it’s the waiters and the cook since the there’s so many analogies of these themes that are suppose to work together but they don’t. You know what I mean?
Male Speaker: Yeah, yeah, yeah, yeah definitely.
Male Speaker: So if you don’t get it, you know, then, you know, then that they, probably not the right guy that have the conversations.
Male Speaker: I think the, clearly this is an opportunity for Andrew Shafer, he needs to have the dev/ops clinic it’s just like inviting over.
Male Speaker: Yeah, he’s got the Internet so far yeah, maybe —
Male Speaker: That’s just right, I saw that.
Male Speaker: You saw — I mean he just, it just kills me man.
Male Speaker: That is good time. So one last thing to mention and is the – this is next it — no, it’s in a few weeks there’s an OpsCamp Boston right John?
Male Speaker: Yeah so OpsCamp Boston when is it? The April 22, Friday. I’m trying to jam up some support for it, you see we are kind of a little low right now, but we’re going to try to get some, bring in some sponsorship to get some free drinks and it’s a good time so “Hey, hey Andy come on down let’s talk dev up.”
Male Speaker: Maybe we can get Andrew to come here, get a boxing ring in there.
Male Speaker: And are you going to, are you guys going to be there Mark like you were at the Austin one?
Mark Hinkle: We’re probably not going to be in Boston.
Male Speaker: There’s a San Francisco one as well.
Male Speaker: Yeah, we’ll be at San Francisco. Yeah, we have gone to the LinuxFest Northwest.
Male Speaker: Yes, that’s a classic for you guys.
Male Speaker: Yeah, so you basically it’s the farther northern point in the continentally class I think. So it’s not easy to get to.
Male Speaker: Alright.
Male Speaker: So, I’m definitely going to be in the Boston one, I’m one of the sponsor or one of the, what do you call administrators or whatever, like I said. So yeah and so I hope, you know, people can, you know, who are in Boston area definitely stop out, there will be some drinks, some food and, you know, like all, you know, if you’re interested in dev/ops or operations in the cloud or just operations in general, it’s a great, I mean, it’s, you know, all you guys, you know, we were all at the Austin one and that was just a blast.
Male Speaker: Yeah, it was, it’s free as well, so they’re free to attend.
Male Speaker: Free.
Male Speaker: So there you go.
Male Speaker: Well, I got to wrap up here and I’ll put some links to these few things we didn’t mention but there’s also like, also what over the past since last episode, there’s actually been a lot of releases of some interesting IT management things. There’s a sell ops which had, which is, they have virtual appliance and they also have a SaaS hosted monitoring thing and they have the new version 1.6 out. And one of the podcast favorites Spiceworks, they have a new version out that’s adding a lot of functionality. And they are interesting to watch because they have, you know, they’re interesting on their own right.
And as I’ve said, before they have ambitions to kind of go a little bit up market so that’s they’re interesting guys to check out, to use that word the 100 time. And also, I think I’m the only one who ever writes about these guys but there’s little Canadian company called Versera that’s has a monitoring that’s a virtual appliance as well and they have a free demo version that’s SaaS hosted you can go checkout, they had another release out. As a very small footnote one interesting thing is they’re actually HTML5 to do a network topology thing instead of Flash, so go check that out kind of interesting. Used the word 101st time. But since I got to run that’s all the time we got for this episode. Thanks for being the guest on this Mark and that we’ve enjoyed the typing.
Male Speaker: It’s Matt, it’s not me.
Male Speaker: I hear you, I hear you. Well, he’s a busy guy. He has got lot to do.
Male Speaker: He is a busy guy.
Male Speaker: Who was that guy that was typing in background?
Male Speaker: Well, like you said —
Male Speaker: I know, I hear, but I can tell by his typing, that’s Matt right?
Male Speaker: That’s right, we’ll see for our 2011 prediction show if not —
Male Speaker: That’s right and with that we’ll see everyone next week.
Disclosure: IBM, Eucalytus, Dell, CA, AceelOps, Spiceworks, Zenoss, and OpsCode are clients. See the RedMonk client list for other relevant clients.
[…] IT Management & Cloud Podcast #69 – John and I go over a lot of BigCo cloud news, like CA buying up a bunch of cloud assets. […]