Skip to content

Own your own stack – IT Management & Cloud Podcast #071

The Old UI & KM - Mainframe Chorus at CA World

In this network connection cursed episode, John and I hit up some recent SaaS and cloud offerings and acquisitions. Also, trying to figure out how companies should take advantage of the cloud.

Due to CA World travel – actually, forgetting to copy the recording onto my laptop – this episode is up a bit late.

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

  • SAP buys Sybase – Coté’s Quick Analysis. We get John’s veteran take, esp. from a Tivoli perspective.
  • VMforce – Coté on the topic.
  • CA World – pre-conference interviews point towards lots of cloud stuff – which ended up being the case.
  • IBM buys Cast Iron Systems – James has two pieces covering it: here and here. then we talk about cloud integration and the brakes being put on complete cloud integration, security FUD, and the comforting alternative of hybrid clouds. Them Pervasive folks in Austin seem to be on this too.
  • Cloud computing nuts gotta address the security fears of mainstream adopters.
  • OpsCamp SF – coming up this [past] Saturday.
  • What’s up with OpSource? From back-office SaaS running and support, to public cloud for the enterprise.
  • Remedy OnDemand – some more info out there.
  • We need to hear from more users of cloud stuff, not just vendors.
  • Own your own stack – John and I blue sky on businesses taking advantage of new technologies to deliver better applications/engagement with their customers.
  • John starts up dev/ops cafe podcast with Damon Edwards.
  • Conferences: Gluecon, Velocity dev/ops day.

Full Transcript

I don’t look over these closely, so if we say something whacky, check the original audio first. Also, where there were transcription issues (garbled audio, etc.), there’s a time-code dropped in so that you can find the original audio to check.

Michael Coté: Hello everybody! It’s the IT management and cloud podcast, episode number 71 on this 13th of May, that’s Thursday not a Friday, 2010. This is one of your — what do I call myself a co-hosts here as always Michael Coté available at, and I am joined by the other co-host.

John Willis: John M. Willis at, but also we founded

Michael Coté: We got — get a little bit of your giggy working in there somehow.

John Willis: Right.

Michael Coté: So I was complaining to John that I have been having networking problems earlier today. So we will see how our little ride here goes. Hopefully, if anything happens to the magic of editing aside from this babbling on here, you won’t notice that anything happened, dear listeners.

So I was noticing, it’s been almost a month since we recorded, John.

John Willis: Oh really?

Michael Coté: I think so.

John Willis: I think we just did one a couple of weeks ago.

Michael Coté: It may have seen like a couple — but according to my records episode 70 was posted on April 16th.

John Willis: I like that, boy!

Michael Coté: Yeah. But we’ve had a lot of exciting adventures between now and then. A lot has happened.

John Willis: Yes. Yes sir.

Michael Coté: And it’s a good question of where to start. What are you feeling over there? We have got all sorts of stuff going on.

John Willis: Yeah, I guess there were some big acquisitions. Do you want to start with some of the acquisitions I guess?

Michael Coté: Well, I’ll tell you what I just spent like an hour or so or more writing up and that was you got the SAP buying Sybase. What can you tell being an old school guy, or a guy who is around during the old school? What’s up with these Sybase cats? What’s their deal, man? It could be all jive.

John Willis: I mean, they aren’t cats, man, they are dudes. It’s weird, I mean — I got my baptism of distributive computing with Sybase really, I mean. When I first started doing Tivoli stuff, everything goes back to Tivoli. You should call the IT management podcast and John always mentions Tivolis.

Michael Coté: We just need a little sponsorship lined up.

John Willis: There you go. Where would those IBM guys could dump us some money, right? Let’s say you think most of the time about that. But the early Tivoli, the original Tivoli before IBM purchased them, they only ran on Sybase. That was all there, well, like their Event Manager and all that stuff, their database or Sybase.

So when I first started installing Tivoli, you had to go through this, the manual Sybase, and I used to swear that it was the worst install process in the history of installers. It was just as like, kind of screen text prompting that just was, god got awful, and if you made one mistake, they blow everything away and start over again.

But the database was almost, I mean I think those guys always had it going and I am just like I never really understood why RAOracle aid their launch, but I am not a database guy. I always look to Sybase database from the functional, and it is like I hated the install process. Once you got it up, I mean it seem to smoke.

Michael Coté: Yeah, now I mean, if you go look at the Wikipedia page of Sybase which is, Wikipedia is always a good source for figuring out the older companies, but Sybase is one of these companies that’s been around forever, and my being a little newer to the scene as it were. My take is, back in the olden days, when I was just a lad or whatever, as I have read there was some sort of relational database war is going on. Yeah, the Oracle and the Sybase, Ingress, and the DB2 people, all sorts of people —

John Willis: All informants.

Michael Coté: There was even like — then there was other weird birds like Paradox and FoxPro and all these things, right? And I think Oracle came out as the clear sort of winner of that with folks like DB2 and others. And I guess IBM bought Informatics. But you know, Sybase was one of the — I guess you can could them losers of the database war, it’s in the sense that they didn’t come out as a dominant player, and yet what with the news of SAP acquiring, I mean, there is plenty of 04:25 or I guess it’s pixels nowadays, plenty of pixels spent on like talking about the revenue that Sybase has. It’s something like, I don’t know what it is, $1-2 billion or something, total revenue that they have and 800 million of that which is database base.

So as you and I are always kind of joking like, nice revenue if you can get it, but clearly they are not — they didn’t emerge like Oracle has from the old —


John Willis: Well, you know the message there, it’s the same thing as IBM, is that the — it just do not get rid of software, you know what I mean, I mean it’s just you know, they keep those license. Once they buy enterprise software, they just never get rid of it, and Sybase sold a shit lower software in the ’80s and I guess late ’80s, during those databases or coming into the era, I guess early ’90s, I don’t know, so they definitely, you know, there were on the do it part, it was almost like Sybase was, I mean, it was kind of like, Sun/IBM in some ways.

Oracle was like IBM and Sybase was like Sun, you know what I mean?

Michael Coté: Right.

John Willis: I found that like there was a lot of Sybase in Europe and you find it a lot of – well, at the enterprise level, you would find more Sybase implementations in Europe. These were from a Tivoli perspective and then more Sun, people going really big arm with the Sun hardware in Europe than they would in the Europe. So, I don’t know.

Michael Coté: That’s funny because the little quick analysis as I like to call them that I just wrote up like I started out by explaining that Sybase is often one of those who were big in X companies, where like, you haven’t really heard about the mainstream but they paddle off all these things. Like you are saying, they were big in Europe.

And apparently, there is some sort of — I have to admit my thorough understanding of Sybase is anything but thorough, but apparently they have — they have a good foot-hold in doing some of the sort of backend, doing SMS stuff for mobile, and there is some mobile stuff they bought recently for relatively recently for handling, I don’t know, doing mobile stuff. They’ve got stuff here and there.

Anyways, I have — we are not really application-centric here and so we don’t want to — I don’t want to bore everyone. I would rather bore people with talking about American Airlines’ Frequent Flyer programs than applications. So you can go, look at the sort of post that I wrote up and there is some good links to other people who — I am usually kind of a glass half full guy, but I was a little glass half on this one. So there is some people who are little rosier than you can read up on it to. So it’s kind of interesting, exciting times.

John Willis: Good times!

Michael Coté: So while we were on the acquisition plane here, or whatever, there was also — I think between now and last time we talked there was this VMforce announcement, right, am I getting my 07:33 lined up right?

John Willis: Yeah, now you got it right. I think that was after our last podcast for sure.

Michael Coté: And that was basically a partnership between VMware, and hold on, Salesforce, and I think VMforce, actually interesting name. I think what’s interesting here is the Salesforce is sort of — it’s not out yet, so it’s just announced and it’s apparently in the fall when there will be more details, and pricing, and all the hoop-law if you will. And it looks like so, you had the force — you were able to use force as a Platform-as-a-Service. So you are able to write plug-ins essentially, and applications and stuff that ran on Salesforce’s cloud. But you had to use their custom Apex language and all sorts of things like this.

So on the one hand, this partnership which brings in a lot of spring in Java development is a way to do more normal Java development in the force area and still get access to all the Salesforce customers. And there is actually some interesting – I don’t know, it’s kind of a stretch because DevOps is a very elastic concept so to speak. But there is some interesting DevOps tooling sort of things going on as far as using a virtualized sort of infrastructure on your Desktop to develop your application, sort of fantasy world, or maybe not fantasy world, push-button deploying it off to VMforce and so forth and so on.

We will see when it finally comes out this fall, but it looks like there is some interesting deployment cycle stuff going on there in addition to whatever other real stuff is happening.

John Willis: Yeah, it’s funny — it’s like 09:05 some full circle on the whole where I was like a year-and-a-half ago or two years ago on some of my ideas about things and kind of like the – you know that with VMware, I mean it was clear 09:16 very frustrated with VMware, kind of like the IBMs enjoying their time, digging to feed into like we are a cloud and like probably not. We are a cloud, no, you are not. I mean, out of your mind, you said, okay, well, you know, they will be when they want to be, you know what I mean?

Michael Coté: Yeah.

John Willis: I mean because the VMware, right. But the thing is in like VMware in a lot of ways, it’s just taken this kind of — whether they’ve done on purpose or on a methodical approach where you are looking some of the acquisitions and again SpringSource, and all of the things they have done to get to this point, man, I hate to be continuing with that.

I mean, think about it, I mean from any space, I mean, it just got so many tentacles in all the right ways, you know what I mean, in other words and again there is always a force excused upon, to recon with in the virtual world, but I mean they are really – you think about where they can go with the SpringSource and all that in the app development and jumping into this kind of force, PassWorld, and then they are just credited by the virtualization. If it’s carried, I was like the company was on my radar.

Michael Coté: And we’ve talked about this is in recent months but there is things to things like this and also CISCO getting in the unified computing and all sorts of other things, and Oracle buying Sun, I mean, it is a –if you are a large established tech vendor, it’s a weird time to be in.

John Willis: Oh yeah, cool!

Michael Coté: You are not quite sure, it’s one of those times, where you can’t be lazy and the way that you put it is like every effort what your fancy name for this cycle is if anything, but every ten years the IT staff and operations people get to do exciting stuff then they go —

John Willis: That’s cicada.

Michael Coté: There you go, it’s the cicada pattern of IT. And on the one hand, it’s time to go like be locust or cicadas or whatever and eat each everything up which is exciting, but then if you are sort of like the tech vendors out there it’s a scrambled time and it’s interesting to see whenever you have a situation like this on the one end, you’ve got the people who are doing their best to convince you that nothing has changed, and all this cloud stuff is a bunch of crap, and why don’t you buy some of those million dollar extra data servers that we have?

And then you’ve got all the other people who are like, dude, cloud is the only thing. It’s kind of like the extreme sort of parting there, and there is simply not that many middle ground.

John Willis: Yeah, well, it’s funny, I mean not this 11:56 like a VMware 11:56 but they position themselves right in the middle. I mean I still — when you talk about HP, when you talk about IBM, and IBM has done this in developer’s cloud and it’s literally interesting, but it just really haven’t pointed up anywhere where you would say, okay, that’s happening. In HP, I don’t see anything coming out in those guys, but VMware has played this kind of fox game that IBM has played.

Like IBM was the first big companies that worked out before anything, and then VMware was like, we are clouds too, we are cloud too, and meanwhile nobody really was a cloud, it was just a stake in the ground and stuff. But VMware is like of patiently filled in gaps really well, you know what I mean. Obviously you can’t discount IBM and their webs you expect and all that but I don’t know, I just — I am kind of high on VMware’s position.

But usually if you are following me as a betting man, don’t take any advice from me, I always lose money in the soccer 13:02.

Michael Coté: Well, speaking of people going cloud or whatever, I am going to see AWorld next week for the first time.

John Willis: Oh wow!

Michael Coté: And that will be a good time. There has been a couple of like pre-conference interviews with people here and they all have to find some, the link to where — it looks like cloud is going to be pretty big on the agenda which people who’ve listened to this — regular listeners would know well since CA has bought 3Tera and 13:30 and some other folks.

John Willis: I think that again here is a company that I would never be talking, I am sort of on that, but here again I mean depending on what they do with these recent technologies and acquisitions, again they are another one that’s like again kind of stuck their foot in the water, and we are clouds, they are just yelling as loud as everybody else and then like VMware they have been scooping up some just really interesting properties.

So which again is not really what IBMist can do, you know of course he won’t say that 14:07 IBM.

Michael Coté: So why don’t we – we should keep on the acquisition train here, which is like, there is a lot of acquisitions going on now. I remember there was a story a few months ago about how all the tech companies had billions of dollars of cash just laying around, and they were going to start buying stuff which is obviously what seems to be happening here. But also I mean I guess it’s in the cloud area it’s also just some data integration, but IBM bought this company called Cast Iron Systems.

John Willis: Oh, yeah that’s huge.

Michael Coté: Definitely cast as a cloud buy if you will.

John Willis: Yeah, well it kind of blows my whole theory. They were not doing anything, that was – it’s another – but again, it’s a weird IBM acquisition, you know what I mean. So I think the Cast Iron is clearly — they have got some chops in the data integration, right, and all that stuff. But I don’t know – yeah, I mean that’s a great acquisition, there is no doubt about that.


Michael Coté: Yeah, I mean basically they do data integrate, I mean the cloud angle right, I mean they do data integration which is getting data from one chunk of software to another, or one system from another and doing the transforming and whatever, and everything that comes with that essentially which is sort of a more boring substrate of computation but very necessary.

John Willis: Yeah, they are the guys that support your ass really.

Michael Coté: That’s right, 15:32 your underwear.

John Willis: 15:33.

Michael Coté: And so I guess they have a lot of hybrid cloud integration stuff going on which I think – I think when you look at the — if we can go a little cloud speculation here for a moment, not like we don’t do that all the time, but it does seem like the brakes have successfully been put on, enterprise is quickly adopting cloud computing in massive ways, which is more or less expected. I mean, I don’t know if there’s ever that many big shifts that happen in 12 or 24 months.

And so it’s interesting like I have been purposefully noticing that again-and-again, I used to just miss this kind of stuff, but what comes up is people, enterprises or companies or whatever are like, hey, I love to do that cloud stuff, but I am worried about security and all this business.

I mean they are all worried about things and then they will throw things like HIPAA at you and PCI and all this compliance stuff. It’s basically at a point where non-technological concerns I guess security is, obviously technological concern. But there is a lot of cultural and process concerns that are holding back cloud stuff. And the answer to that is to go to the sort of hybrid approach where you can stay with the old and you can go with the new, like all that blue stuff you are suppose to wear at a wedding or whatever.

And I think things like Cast Iron are like they fit in excellently with that strategy, if you want to allow people to move strategically, and what I mean by strategically is they choose low-risk stuff, kind of integrate stuff with the cloud but keep their on-premise things.

There is like Identity Management company here in Austin called Conformity, that’s another interesting approach of doing this, and I think even the speaking of the relational database guys, another place outfit here in Austin pervasive, I think it wasn’t data general but it was data function they bought a long time ago. It does a similar, a lot of cloud integration.

John Willis: I met those guys and so I came in early, I am going to OpsCamp 17:45 OpsCamp in San Francisco. I think we’ve had over 150 registered, so I think it’s going to be like Austinish, so have that kind of, I mean —

Michael Coté: Is that in Boston or where is it?

John Willis: No, it’s in San Francisco.

Michael Coté: Ah, that’s why you are in San Francisco.

John Willis: Yeah. So I came in early, there were all these cloud things going. Citrix Synergy and there was this OpSource and then there is a service company that we are going to partner with and so these guys are all with the OpSource thing. So, I felt, you know what the heck, I will just come in on Wednesday, meet with all those guys and stick my head on these various conferences and see what’s going on.

So going up to the OpSource cloud, it’s the SAS cloud, or your head in the cloud, your ass in the cloud, I don’t know. So I happened to get in there, and it was the end of the conference I didn’t know. I met with these guys and then I went over, it was the end of the conference. So I just asked the people for registration where I said them, I just met somebody for a meeting, I am just sticking around. Do you mind if I just go into the demo session? They were like, yes, go ahead, so they just let me go in.

So I got to hear the last two sessions and it was kind of funny, the second last session was a panel of experts on private clouds, and then the last session was a panel of experts on public clouds. And so you are talking about all that. So it was like so I listened to all the guys on the private cloud, it’s like they all have their spin. The private clouds are real kind of a 19:07, no spin here, they are very – I don’t know — and it is like when people get religious about things, they blurred out things, and they kind of ripe but they are also like they are all you would be an idiot if you didn’t understand why you need to do this.

And so the next session they had like a bunch of public cloud guys and they were the same way, they were like just blurring out like, you are an idiot if you think security is a concern in a cloud. There is only — no private company could ever get the economy as a scale that a cloud company can get, you know what I mean, the guy who is saying is like runs a small little –it’s OpSource right, so he runs and OpSource’s cloud. And he is talking about how no private company can get, an OpSource can’t be more than, I don’t know how much money they have got, and it’s like, wait a minute, dude, you are trying to tell me the JPMorgan Chase can’t get the economy’s scale for infrastructure that you guys can get? Come on!


They probably own most of your notes. You know what I mean? So I just — so I got up at the end, you know me — I had — even though I wasn’t interfacing or anything, I get up, they give me the mic and I am like, alright, here, we are going to go.

I just took them all to task. I was like, one after another, I made a note about three of them. You know, it’s funny, I am a fan of public cloud, so don’t get me wrong here, but it’s funny to hear the spin difference between the private cloud guys and then you guys’ spin and any old chuckled, and I just nailed in to them. I nailed it into the guy from OpSource and said, how do you say that private enterprises can never get the economies at scaled that 20:48 or GoGrid gets? I mean, that’s silly.

I mean, they can build an infrastructure like — in fact, Amazon, if you look at Amazon, there is all these theories about how Amazon got into the cloud, but I mean, I have heard people at conferences say to me that, one of the reasons they got into it is because they realize they had so much buying power, that it was a natural fit.

And I think it is companies like Exxon and JPMorgan Chase and Bank of America, I mean they get to buy equipment at prices that those guys will never get to buy.

Honestly, Google gets to buy probably — but I mean, some of the — again — anyway, but the point is to say that these guys can’t build economy of scale infrastructure, when JPMorgan Chase, and I have said this a hundred times on this show, is 17 operating businesses. It’s silly.

I was trying to come up with a top ten list of private cloud spin and then a top ten of public cloud spin.

Michael Coté: Oh, that’s fantastic.

John Willis: Isn’t that good? Yeah. I have only got like four on each category so far, but we can slice through that. I mean, the whole public cloud is, you are an idiot if you think security isn’t into the public cloud. I mean, that’s how blatant it is.

Particularly the Amazon guys, they will sometimes like, it’s just puppy talk that people think that public cloud is not secure. Let me tell you why it is.

So when I was nailing those guys, I am like, it’s secure.

Michael Coté: Hey, we are back.

John Willis: Like I was saying. So I was just trying to — the whole private cloud, I have probably said it three times now, but saying private cloud doesn’t exist, being a public cloud vendors —

Michael Coté: Like I said, I mean, I think the upshot of that is, is I think the early adopters of cloud computing like figured out that security — this happens with every technology. Like my — security always becomes an issue when you want to go to mainstream or wider adoption, and people always worried about that. And it’s very frustrating if you are one of the people in the community that’s been using it, because you know it’s secured, you believe in it, and then you also have those devilish arguments, where you are like, well, how secure are you guy?

John Willis: Yeah, yeah, yeah.

Michael Coté: Which are sort of like rhetorically valid, but persuasively they are not very affective. And so it’s kind of — I think people who are doing cloud stuff, it’s kind of incumbent on them to speak to the security, whenever they do, because — I mean, I will tell you what, that’s like the number one thing I come across time and time again about people who are mainstream and they don’t want to use cloud stuff. They are like, well, is it secure? Which is fine. So you have got to address that need.

John Willis: Well, yeah, I think those guys stand up and say — trying to make everybody feel like they are an idiot, because they are like, well, I heard it might not be. Oh, you idiot. The cloud is secure enough. You know what I mean? In other words, for most things, yeah, it probably is secure, but last year I had this discussion with this guy who was on the CIO team for the 24:03 events. And I was trying to talk him into — or trying to broker a deal with him — not broker a deal, I was trying to recommend some solutions, and one of them was CloudSwitch.

So we talked about CloudSwitch and he asked me a whole bunch of detail. CloudSwitch is the one that they kind of pull in the VM. They encrypt it. They run it in kind of a micro hypervisor, what they call an isolation technology. It’s pretty cool, but he was like, well, what is — is it encrypted in memory? And I am like, no. There is certain things we just can’t put out.

This idea — there were guys in the military that if you take a 24:40 offer something, and that data is not encrypted, or you get a memory snapshot or a running snapshot of data being processed, the question is, what is the certifications of the people that might be able to see that data?

Michael Coté: Yeah.


John Willis: In the military, it’s people to have top secret clearance or higher, depending on the data classification of the data. So military is the extreme case, but until you can resolve that problem, it’s not just, well, we have VPC and Virtual Product Clouds and all this stuff, no, I mean they have some real issues about who gets to see and has access to that data.

And I think Amazon would tell you, well, you know, our guys go through — the people we have, our operations group have been through SAS 70 Audits and all of this blah, blah, blah and blah, blah, blah, but the guys who get to see data — so for example, the guys who get to see data at Bank of America, have phenomenal background checks. You know what I mean?

Their financials have been, if you are going to work — if you are going to be in operations at Bank of America or Federal Reserve, they are going to do like not a credit check on. They are going to basically exhaust everything about your financial history.

Michael Coté: Make sure you haven’t ever a spouse believe in the communist system?

John Willis: Yes right, but the point being that again, Amazon may say, well, our operations guys or we do credit checks and all that. Well, do they do the kind of checks that Bank of America requires for their operations? Maybe not, you know what I mean? So they go around and just make it look like you are an idiot because — and then there is this other — I am kind of in a rant mode but — and I apologize to everybody, but there’s this kind of —

Michael Coté: It’s because we made an East Coast person go to the West Coast.

John Willis: That’s exactly that.

Michael Coté: It’s everything (Voice Overlap).

John Willis: You know it.

Michael Coté: Three time zones later, everything is —

John Willis: Oh yeah, I tested everybody, I am getting in fights with cable car guys, I mean it’s just crazy.

Michael Coté: You like cable cars, give me a car.

John Willis: That’s like, hey, get out of the street, you jerk! But so that’s when I come out here in California. No, but the other thing that they always say is like, well, we find that most of our datacenters are actually more secure than most of the enterprise, and that maybe true. But I remember going to visit there in Annex (ph), that looked like a freaking bunker in a military — I mean it was — Annex has a datacenter in Arizona or they used to, they probably still do. I mean that thing was like tested for nuclear – I don’t know, maybe not nuclear, but I mean, it was a bond shelter.

So when these guys use these blankets, well, probably the chances are that our datacenter is most secure than most enterprise. Again, that maybe true but to try to make it a blanket statement, we are more secure and that’s the kind of translation. So I don’t know.

Michael Coté: Yeah, well, they just didn’t remember that they need to prove these things and not just statement. I mean that’s —

John Willis: Right, and then the other thing like they tend to just gloss over is the multi-tendency, it’s like, we’ve been doing this stuff for 30 years. Well, no, you really haven’t had — I mean hosting yes, but to a certain extent, but had really mixed shared customer data. I mean you take Amazon’s point; there’s really been nothing like that where two customers, Bank of America and JPMorgan Chase may actually run on the same physical server under the same hypervisor. There is no guarantee in Amazon that that won’t happen.

In fact, there is a white-paper by some university that has showed through the actual naming of the InstantIDs and all of that, how it is potentially that Amazon could actually be gained for what Chris Hoff calls a potential Economic Denial of Service.

Michael Coté: That sounds hot.

John Willis: It is hot, that’s pretty cool. But no, Chris came up beaker, he came up with this idea of Economic Denial of Service. Like, can I like really — can I kind of figure out where you are on in your cloud because I am in that cloud too, and then really just make — just drive up your prices. So it was kind of all in theory like could this happen and then actually some universities do the study basically where they prove based on some of the naming conventions, the InstantIDs and all of those things that you could actually — if you use some types of algorithms, you could actually isolate or X is trying to find Y and you can find Y and actually cause some disruption to Y.

Michael Coté: So like you said you are over there in the San Francisco, what’s up with these OpSource guys, give us like a little overview of what their story is?

John Willis: Well, so OpSource started out as kind of a tool chain for So a few years ago, their whole original was they did a lot of the back-office stuff, kind of like your buddies, and you have the company and the ex-BMC guys that had — they were doing back-office for Software-as-a-Service. We talk to those guy; one of them is your friend.

Michael Coté: Oh, yeah, yeah, eVapt.

Guest: Yeah, yeah, so OpSource was one of the big players originally in the back office SaaS space. It’s dual junky back office crap that we don’t want to do and it kind of built there, kind of a tool offer, it became a really powerful business partner/provider around the force in space for a lot of that back office stuff.

Did very well as group, and I don’t know about a year ago, they kind of announced that they were getting into the public cloud space. So they kind of transitioned or re-factored their model to be a public cloud. So now they are public cloud and what they position themselves is, they do a public cloud catering to the enterprise. And the two things that they brought to the table which should made them reasonably successful is, one, I mean, one, I will give them credit, they are actually pretty smart guys, I have talked to their CTO, John Rowell is their CTO. He has definitely got it going.

I think they made a reasonable amount of money, may be a lot of money, also what they had done prior through the SaaS boom and they have run — and they have really used more of a hosting model for a cloud, in other words, they used VMware and the VMware infrastructure, which suits the enterprise. So they are kind of one of these, like yeah, we are a cloud, we operate externally and look like a cloud externally, but we are really just really well-designed secret source scripting VMware implementation.

Michael Coté: I think the SaaS based stuff at BMC, I used to work on Patrol Express, was run by OpSource at some point.

Guest: Oh really?

Michael Coté: Given over to them to run.

Guest: Oh, that’s interesting. Well, today the thing is I think what gave them — they are kind of why they are on the map and they are kind of an important player now, I think they were able to branch back that whole, this new story through all of their customers. So they already have a built-in – there’s a pretty good crowd of people that are willing to listen to them, that they do very well, and trusted them, and two, they have reasonable amount of money now where they are having their own conferences.

So they have their own conference here in San Francisco. It was a cloud conference, it was their conference so they had vendors and exhibitors. So they are on the radar, you don’t know — I would say that in my mind, I see more about them than I see GoGrid. I mean from the US space, they have public clouds, they are probably – I’ll probably get this wrong if somebody was screaming all over at me, but they are probably third or fourth, may be tied with GoGrid or a little higher 33:00

Michael Coté: Terrible! We got some Mickey Mouse connectivity over here, but I got the recording back on. So it’s –. Anyway, yes, there you go, OpSource.

Guest: OpSource, there you go. Good guys, good guys.

Michael Coté: We’ll keep an eye. So bridging that cloud and good old-fashioned IT management. I think also may be between now and last time we recorded, there was a lot more information about the Remedy OnDemand out there, which is, of course, the hosted service desk from BMC. There is actually some — I am not sure if there is pricing and stuff that’s out there yet, but there is a couple of interesting demo videos out there and it looks remarkably like existing remedy stuff I have seen kind of put up into a SaaS offering.

Guest: That must be just unbelievably exciting.

Michael Coté: Now, I am like a service desk to get John.

Guest: Good old product to spend around for about 20 years service desk as a SaaS. This is my point, Sir Coté. That was like that Cloud Expo I went to. I don’t if we talked about that or not, but I mean it was just like — I mean I don’t want to 34:23 expo guys because they were pretty nice to me; they let me come in and speak. They were extremely great but this is just, it was like that’s just listening to just large enterprise type vendors come in and try to just consume the cloud conversation like, we have doing this forever and this is — oh, yeah we get it.

Michael Coté: We are kind of at that point with cloud computing where there needs to be a lot more of users talking about it than vendors right.


And the more — not that there is anything completely wrong with vendors; they do a lot of good work as — hey we’re back again — anyway, I was saying, it’s at this point, vendors are great at the beginning to have a lot of conversation when innovation is going on and they need to explain what’s happening and kind of get involved in the conversation, but at this point, I mean, we really need to in addition hear from more users and —

John Willis: Yeah, well, now it’s offered to vendors. I know I’m in tumble.

Michael Coté: So, whenever — and I speak to this all the time like I really — in the limited capacity, I have to try to call the stuff up, like I’m really interested in seeing panels and presentation and discussions from people who are actually using the cloud and kind of tricks on that. I’m trying to get those folks I mentioned about the emerging technologies and the enterprise comments. There were users who had some really good comments about that.

It looks like at the end of the month, we’ll have them on the podcast, but — that’s a good thing to look out for is like if you’re going to go to a panel, like, try to make sure that there are sort of users speaking to what they’re doing, because it’s going to be kind of the same thing you’ll hear from vendors nowadays, but it’s good to hear what the offerings are, to hear the technologist think, but you’ve got to hear how people are using it and what’s going on there.

John Willis: It’s interesting! I get to in my kind of little spectrum, before I had a job for a living, I could go out and make wild speculations about anything. It didn’t really matter if I made an idiot out of myself or whatever. So now to a certain extent, I could still do a little of that, not as much. But one of the things, I’ve been really thinking about — I don’t know, it can take up an hour here but I had a great conversation this morning with somebody. I was just in Toronto CloudCamp, I don’t know — it was like over a month ago or so.

So, one of the guys there was kind of arguing with me after my session about how I was wrong and operations’ code and does it– he called me this morning like, you were right. I really have been digging in and he was just like light bulb went off like a month-and-a-half later. So we had a great conversation. So, part of what I’ve been saying on this podcast, I was saying a lot is, this is kind of — I think there is this — I wouldn’t bank on it, but I think there is this opportunity that maybe I’ll go back to say, maybe Nick Carr was right, you know what I mean?

That is that when you think about what’s happening with abstraction of the infrastructure, you just find more and more that subject matter experts can own their own stack. I’m seeing more and more of this that it’s part of kind of DevOps but it’s like people can own — you can be in a business, like a marketing or a media company and you can today own your whole stack. You know what I mean, from the development, from the infrastructure to the provisioning configuration management, all the way to the end, and manage it reasonably, which is three years ago, you just couldn’t do that.

I mean four or five years ago, you just couldn’t get infrastructure. Three years, you could start getting cloud like a virtualized infrastructure, which still cost you ridiculous amount of money to have the operational experience. Then it was kind of agile operations or Agile IT. So the other day, so I was talking to this guy, and I was telling him about a story I heard the other day. So we’ve got a customer called Tom Dysinger (ph). He’s got a company called Sony and right now I don’t know exactly what they do, but what’s really freaking cool about what blew me away what he’s doing is — so he is like tied his sprints to his infrastructure and like okay, well that’s DevOps.

So at the end of every sprint cycle, he basically not only comes out with the whole agile structure of his new code and get in deployed quick, he actually rebuilds — he uses a 39:04 infrastructure. So like, in other words, at the end of every sprint cycle, he is actually rebuilding the whole IT infrastructure architecture. I’m thinking, God, I was telling Damon Edwards of DTL about this and he’s like, one of the guy who works for him is, he used to be like the Head of Operations for E-Trade and he said it’s funny that he was like, wow, that is cool!

He said, but do you know, Lee Thomson who used to work for E-Trade, he said that like three to five years ago, he suggested that at E-Trade, every morning, they should basically refracted the infrastructure before trading. They were like they laughed them out of the building, you know what I mean?

Michael Coté: Yeah.

John Willis: I’m like — so I’m getting to a point. I think about people owing their own stack. I think more about that if we really wanted to be agile, we know we want the whole soup like the whole enchilada to be agile.


In other words, if we are going to do a sprint, and we are going to get together a sprint and we are going to go through this kind of put it in development, and put it in product, I mean put it in test or the QA, put it in development — oh I keep saying it, in production sorry, why don’t we want to do the same thing with our infrastructure?

And guess what fans, we have the capability to do that with tools like Cfengine, Puppet and Chef to be able to create disposable infrastructure. We have got tools that do continuous deployment. So the more and more I think about this, I think that mainly it’s going to be very difficult for organizations to own their own IT. And then you say, well, that’s the original cloud argument that everybody is going to go and is only going to be five big companies running IT infrastructure. So I think that Nick Carr may have gotten it right but for the wrong reasons.

Am I making sense?

Michael Coté: Yeah, yeah. I mean this is something that I have been around when you talked about recently that is – like to put it another way, there is this interesting sort of paradoxical thing where like you are talking about outsourcing; other people being more effective earlier.

John Willis: That’s right.

Michael Coté: And to some extent, I am sure there are some better empirical studies than me. I am still in the role where I could just crazy things. So I will go ahead and do that. It seems like —

John Willis: Well, isn’t that the point of our podcast?

Michael Coté: Exactly! It seems like the benefits of outsourcing IT have been mixed over the past 10 years or so, like it’s — sometimes that works and sometimes that’s utter failure and all sorts of things like that. And to some extent you have got this game of focusing on what’s valuable to you as a company, and then outsourcing what’s not valuable.

John Willis: Right, right.

Michael Coté: And it seems like — it would seem like in most situations, the application level, everything at the SaaS level is valuable, but everything below it is perhaps not really too valuable, and it seems like maybe with the cloud stuff and all this hoopla that we are always excited about, there is a chance to outsource or somehow minimize the amount of time and resources you have to spend on the lower levels which frees up time for you to spend on the higher levels on application development; actually, doing something of business value with your IT other than just delivering the email and complying to Sarbanes–Oxley and crap like that.

And so, I mean that’s like when I start to talk with more people about like what the hell do I do with this cloud stuff? I mean I start to sound like old BMC, Business Service Manager and Messaging or whatever, but then the top level messaging, I don’t know what that means, top level messaging, but the goal of all that really was pretty nice.

It was sort of like wouldn’t you rather not worry about all this crap, instead focus on applications? And I think what you are getting at with this owning your own stack is sort of like the tactical, strategic stuff that that vision amounts to is that if you are not writing your own software, then something is probably going wrong.

Like if you are just customizing someone else’s software or you are not — companies that write a lot of their own software, not all of it. I am not suggesting that you shouldn’t go out and buy like your ERP systems, because it’s really dumb not to – you want to get off the shelf for counting stuff.

John Willis: Right, right, right.

Michael Coté: But the closer, whatever software you have that’s actually touching the core of your business, the closer to the money or the customer that you are, it’s probably worth considering if you should be writing that software on your own.

And then even more importantly, making sure that you are actually writing a really good software there. I mean anyone — there are any number and these are all consumer examples I will go through, but there’s any number of online banking and like AT&T, like login stuff, all sorts of software that I deal with as a consumer from giant enterprise companies.

Nowadays, it usually looks good, but it’s not really very good software, like it’s not always as good as it could be. If they would just take some of those millions of dollars spent on infrastructure management and like spend it on better user, just better user experience stuff, then I don’t know, the world would be a better place in the little way that IT can effect it.

Hey! We are back now. You know what I am going to do, John, and this is my gift to the listeners. I am going to pretty soon here, maybe tomorrow, I am going to go across the street, we have got a clear wireless outlet, I am just going to buy myself my own Internet. I think that’s going to work out better.

John Willis: There you go.

Michael Coté: Because whatever I have got going on here is not working. But anyways, I mean I think the frothy consultant speak vision crap that I was getting to is just basically like – I mean really, I think companies have a limited budget and attention span and everything. And if we can get them spending less time and money on infrastructure crap and take that money and instead spend it on like just making their actual applications better. I think –


John Willis: Yeah, I agree. I mean I think that’s the – I’d say, you own your own stack, but it’s here kind of full circle. It’s use SaaS, I mean whatever way it comes out, whatever SaaS will look like in five years ago for now to the menu, if you will, of enterprise because the enterprise has always tried to be for as long as I have been involved in it, they have always tried the whole of the gate or guard the castle of, we have to do IT for everything we do, because we can do it better and we know all the critical points.

And then there is a lot of truth to that but you know that — I mean if you look at the explosion or if you try to track that how much technology change from the first mainframe to two years ago and what has happened within the last year-and-a-half, and what’s going to happen in next five years, it is what I call like Cambrian Explosion.

So I think that 46:18 IT has always run into massive changes that caused their young ones to run away but then they somehow get it back. I mean if you look at the mainframe to distribute it and then they kind of work their way into, okay we got it all back and again, it’s my job to about every 16 years, 46:40 guys get to have fun. And then the Internet kind of let everybody spray off a little way and a lot of enterprises have got — banks have got their own Internet banking and whatnot. But I think this time, they are not going to be able to catch all the young ones, I mean, because the technology is just —

Michael Coté: I think we have both mentioned it but I think I mean Internet banking is an excellent example of this kind of thing where it’s — or it’s an affect of a similar thing they used to happen where someone is like, Holy Crap! The Internet and they figured out how to take advantage of the Internet to actually do something new and interesting.

John Willis: Yeah, yeah.

Michael Coté: Now, I think people have forgotten that you used to actually have to go into a bank to do much of anything. You could go to an ATM to get money and deposit checks, but that was about it, whereas nowadays just having bill pay is phenomenal. I think we take that kind of stuff for granted and I think there is similar sort of like — I don’t know I just don’t feel like we have had much of a jump in anything interesting in the business use of IT since online banking and online retail.

John Willis: So, here is another fundamental change that — so there is one fundamental change that has to happen or I don’t know what happens if doesn’t. In other words, people are going to have to in the 47:57 understand that there are changes that are just not going to allow that. I definitely could be wrong but I don’t see it happen where they are going to be able to be in the IT one-size-fits-all for this explosion of technology where you have to own your own stack solutions. I mean your marketing department is going out to people who can rapidly deploy scalable Facebook option or I mean just almost every silo in the enterprise, where the enterprise try to — and I know I am beating a dead horse here, but — every silo that the enterprise —

Michael Coté: We just can’t win with this recording here. Hey there!

John Willis: Hey there! This is John M Willis from So I was saying, my point is I think we have beaten the hell out of this is that this is a fundamental change that the IT 49:01 people are going to have to deal with and realize they can’t be the gatekeeper for all of IT activities. But there is a second fundamental change that I think is just as important that could be the death of just IT infrastructure operations as well.

I mean yeah, I know people take on operations here or there are some people out there, but I love 49:23 and I think it’s the coolest thing on the planet. But there is this kind of like I see this chasm of people, like the other day I Twitted that designing and expecting failure is the new black.

It is funny, most people that follow me are like cool or just re-Tweeting it or adding some pretty comments around it but then every once in a while I get a comment from somebody like what the hell are you talking about and that’s crazy speak. It’s almost like I talked at the Cloud Expo, there was that Tweet where the guy said, well it’s the Twitter.


Paramount says server failures are a problem and Amazon says they aren’t. So, I think this is another leap that enterprise people are going to have to learn a lot fast, and that is, you are going to re-factor the way you think about architectures in this new world.

I mean, we have talked about this before, design for failure, build the architectures that are more message bus or the concept of a server really does this. There is no coupling or decoupled architecture. There is no coupling to a server. The idea of a server failure is like who cares, I have got — that’s a worker bee. It’s like the million ants that carry like a rotten apple.

Michael Coté: You want a robust enterprise grade stuff.

John Willis: Right. It’s like if you step —

Michael Coté: What you really want is a flexible infrastructure that isn’t –

John Willis: Right, in other words, my ant analogy was I don’t do this now, but I was a little kid, I stepped on a lot of ants. But you stepped on a whole pile of ants and they still got that apple over there. They were clustered pretty well. So I don’t know. I think that’s – I don’t know. I think a lot of that is, how is the enterprise really going to cope with this.

In some ways I think the enterprise has got — they have got an apple, and I know there’s some enterprise that will never go away, but if you think about the two major, what would you just say, what would you call it, two major kind of changes that are facing them. One is to learn to embrace 51:42 and also learn– because it trains so hard from the IBMs and the HPs and that failure is a problem. And the way you fix it is that you buy a bigger machine and the more money you pay for that machine, the more secure you will get, and the more money you pay for your software, the more safer you are going to be.

If you always buy IBM, you will always be safe. If you buy the most expensive hardware with all the bells and whistles, and you buy this like ridiculous geographically based billion dollar DR redundant systems, you are going to be safe and guess what. I was with a healthcare company where they spent like some ungodly amount of money to bridge their two datacenters that were across the street, and the first time it went out, it was an abject failure. Nothing worked.
So guess what folks, and I love what Jesse says failure freaking happens. So if you think you are going to beat failure, you have already lost.

Michael Coté: That’s right. Well, speaking of not willing to beat failure, why don’t we wrap this one up a little early before we could –?

John Willis: Yeah, a little close to that. Well, –

Michael Coté: And ironically for all of our talk about the cloud, 52:58 activity.

John Willis: Yeah, like alright folks, listen here, do as we say not as we do.

Michael Coté: That’s right, that’s right. And there is a couple of other little links that I will put in the show notes, like Heroku got another series of funding and there are some interesting things they were talking about like going kind of implying about going beyond Ruby as the main runtime for the Platform as a Service that they have.

And there is a – Netflix has a good case about using the Amazon infrastructure to run a lot of Netflix’s operation which is an excellent example of like what John and I were just talking about. Now granted, it’s consumer centric but I would assume that it aligns with what we were talking about because Netflix can — they can focus on the actual application they are delivering; videos-on-demand and whatnot, rather than having your mail, rather on focusing on managing the infrastructure. And there’s a couple of other.

I was also on the ITSM Weekly Podcast Episode 13 as a guest which — it’s a fun little podcast. So it’s worth subscribing to if you like this one, and you started up a new podcast.

John Willis: Yeah, DevOps Cafe with Damon Edwards, yeah. So our first one was pretty cool. We have got this guy from Australia who is – he is nagios. The nagios-cucumber or cucumber-nagios, I like the way those guys say nagios. But yeah, it was very cool. So it was fun. So we are going to just focus on just getting all these to DevOps to try and explain what they think DevOps is and so that’s cool.

Michael Coté: Oh that would be great. Yeah, I was thinking the other day, I Twittered about this that I need to write like a little four-page definitional paper, not that I am trying to put a stake in the ground or anything, but I keep wanting somebody to point to about what DevOps is.

So my plan is to theorize on my own and find something. More importantly, I want to like send out a little survey and see what people say. But I mean if you guys, when it comes to profiling, people are doing that I think that’s always a –

John Willis: Yeah, well, we have got you on the list; we have you on too.


So we want to have you on probably – I think we’re going to try get John Allspaw next, but I think you might be next on the list. I think to get you on, just — we just want to hear everybody that’s around the space. We’re not necessary saying, people are coming on, you know like Allspaw, he is an expert, but we just want to get here everybody’s kind of opinion of what you think DevOps is.

Michael Coté: There you go. We’ll put a link to that in the show notes.

John Willis: Yeah.

Michael Coté: And with that hopefully, by next week, I’ll clear up all this networking stuff.

John Willis: Yeah.

Michael Coté: We’ll see how that crap works.

John Willis: Hey, echo, echo. So 55:35 so we’re going to be there. That’s going to be kind of fun. And then DevOps day, Velocity/DevOps day is shaking up to be like — it’s going to be a rock n’ roll part in that.

Michael Coté: Oh yeah, I need to —

John Willis: You 55:50 DevOps day.

Michael Coté: — figure it out.

John Willis: Yeah, you’ve got to get there. I mean it’s like great panels and I mean this is going to be – really, I mean some of the guys are heavy hitters; 56:00 Randy Bias, John Allspaw. Luke Kanies is going to be on the panel with 56:06 Adam and I have got the guy who wrote RPM, who is this Erick guy from 56:10 and actually 56:14 if you can’t get Mark Burgess too so. Yeah, so it’s — I don’t know if we get other guy but it’s going to be a rocking, rocking show. DevOps US is the place to be.

Michael Coté: There we go. Well, that sounds good. Well, we’ll see everyone next week.

John Willis: Unless you want to hear my crazy story about sending a letter to the President of Delta, but we’ll save that for the next time.

Michael Coté: Alright, alright, let’s hear it.

John Willis: So everybody, so bye to everybody who doesn’t want to hear wacky airline stuff and I know Michael loves this kind of s**t. So I was like furious on my flight. You know like you, when you get to start well known with your airline, we kind of get comfortable with the idea of getting upgraded.

Michael Coté: Yeah.

John Willis: It’s always the worst freaking flight where you don’t get, where it’s like packing of sardines, and there is like no good 57:05. Even though you have got, got your seats 57:08. So that’s what happened to me on this flight. It was like one of these ridiculously packed 57:13. So I just get steaming about I fly too many miles with Delta to not get upgraded. Who the hell got upgraded with me and I’m going to make platinum in June and this is crazy.

Now I am just getting all of that, and I am thinking you know what, one thing you’ve got to do is they don’t have a bidding system for those. You know like kind of eAuction type thing where, wouldn’t it be cool like if you are like in the Delta, you have this Gold, Platinum, and Diamond. So it was like, wouldn’t it be cool if your Gold, you got kind of zero dollars for your bid and if you were Platinum, you got 25 and if you were Diamond, you had 50.

So I was telling the flight attendant about it, that’s a great idea. So you could do just you’re Diamond, you just automatically have 50 in 58:01.

Michael Coté: Yeah. So if you’re Diamond —

John Willis: So listen me off. So, if you are Diamond, you get $75. Did you get that part already?

Michael Coté: Oh yeah.

John Willis: Yeah, right. So then everybody could bid. So if you’re Diamond, you already have $75 in as a bid, but if you’re Gold, you could actually bid you know — I am sorry, if you’re Diamond, you have $50 in, but from Gold and I really want 58:31 last minute bid 75. You know what I mean or you know to bid out some of those guys. You can gain the system and the airline make a ton more money and at least, here sometimes when you show up, you are like 35th on the waiting list. You know like —

Michael Coté: Yeah, yeah.

John Willis: Just feel like crap. At least, if you knew you could bid for — you know like on a red eye. I don’t know, the flight attendant, so I said — she said, well, send a letter to the President of Delta. So I sent a letter to President of Delta and I asked him, this is the part you would really like. I said, if you like this idea and you decide to use it, all I am asking for is a lifetime Diamond membership on a —

Michael Coté: Nice.

John Willis: There you go.

Michael Coté: That’s right. And has he replied yet?

John Willis: No. No, he is never going to reply, but —

Michael Coté: You should publish it on your, an open blog to the open — what did they call that letter, that’s what they called, an open letter.

John Willis: Yeah, I should – yeah, that’s a good idea. Absolutely, you are the smart guy here.

Michael Coté: Well, you enjoy your time in San Francisco then in OpsCamp.

John Willis: Yeah, so well.

Michael Coté: If you come across any people who are actually using the cloud who are vendors.

John Willis: Exactly. Hey! Let me tell you about the cloud.

Michael Coté: You have got about your magic recording box and document it.

John Willis: Yeah, yeah. Well there will be some good stuff from the OpsCamp for sure.

Michael Coté: Definitely. I think there will be a good crowd there for OpsCamp. That will be interesting stuff.

John Willis: Yeah, yeah, good stuff.

Michael Coté: Alright, well once again, we’ll see you everyone next week. Thanks for putting up with our crappy connection.

John Willis: Yeah.

Disclosure: SAP, Salesforce, IBM, and John’s employer OpsCode are RedMonk clients. See the RedMonk client list for other relevant clients.

Categories: Cloud, IT Management Podcast.