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:

This week, John and I meander around several topics:

• Baltimore fun facts: umbrellas.
• Google building a data center in Belgium.
• New restaurant update: Frank in Austin ("hot dogs, cold beer"), glazed donut burgers in Atlanta (like this?).
• Hadoop, The Definitive Guide - a pre-reading book review.
• The "Lean IT" meme - from CA, or Forrester? See a FAQ here.
• And this leads me to ask: has the phrase "BSM" and "business service delivery" talk flipped the bozo bit for marketing talk? John remembers some tales in this area.
• John tells us The Good News about BSM, tough-love version. You've got have your plumbing fixed before applying "BSM in a box," but how well does "fix your plumbing in a box" sell?
• John's review of a recent Gartner webcast on CIO priorities and performance.
• John's DMTF wanderings - OVM, OVF, vSphere deployment. Maybe Winston Bumpus. XML templates for virtual appliances.
• Dealing with PR email, embargo's, etc. For bed time reading, see the infamous TechCrunch take on embargoes.
• Rivermuse launchy - event management, correlation, etc. What is this "event management" anyhow? John says SMARTS is dandy here.
• Our man William over at Oracle has been doing nice stuff on his blog of late, lots of details on how all this IT Management gorp should be designed. I esp. like this line from one post in his series on REST in IT Management: "I can think of ways in which some REST principles would help in this area, but they are mainly along the lines of 'any consistent set of principles would help' rather than anything specific to REST." Yuh! Sidenote: (a.) "REST" is an idea like "Democracy" or "Christianity," good luck getting any agreement on what it actually looks like in practice, and (b.) better luck having the canonical use be practical.
  
• memcache meme - Gear6, North Scale, so on and so forth. Seems like it's on the open source enterprise creep-in path.
  
• IBM buying SPSS - John's SAS memories - also see James Governor's take on the buy.
• Tech books - what do we have?
• New T-shirt slogan: "there's a cloud for that." re: Rackspace private cloud.
• Tell us what books and fun facts and cities you like on the #ITMguys.

Disclosure: many folks mentioned are clients, see the RedMonk clients list for which ones.

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:

John and I review the week's IT management cloud news:

• Lunch with Doug McClure. What's going on in the BSM world now-a-days? IBM BSM stuff.
• What's the BMC offering? "Working with BMC, IT organizations will be able to extend their internal datacenters to Amazon EC2 via a unified, integrated BSM management solution. Enterprise customers can request computing resources – either internal (physical or virtual) or external from Amazon EC2 – through an integrated self-service portal. Those requests are tracked through BMC’s robust ITIL®-compliant change management system and automatically provisioned and configured in minutes. This self-service interface also supports service de-provisioning and service change requests." John says this seems like a better start than what IBM has been doing (or not) in this area.
  
• IBM Tivoli partner conference - selling a lot of modeling and event correlation stuff, so partners speak to that. What types of partners: government with identity management and other area, storage is classically a strong partner, energy management.
• Tivoli foundations products - virtual appliances with full-stack. Pre-integrated Tivoli stack.
• Rackspace's API announcements (first, getting beta ones, then releasing the spec) - having them and then Creative Commons'ing the spec. As John says, Amazon is Apple and Rackspace is Google AppEngine.
• Maybe PaaS is about easier deploying for developers - EngineYard and Heroku. "All I want is to take a WAR and deploy it": we need a PaaS for Java, where are they?
• "Skip cloud, go right to Hadoop" - John's experiments with Hadoop and IT Management performance metrics, 6 months worth. It's easy to add new, unstructured data to existing data sets (a smart, new insight). For example, tracking email campaigns for effectiveness based on state, etc.
• All of the Hadoop examples I can think of are something along the lines of "retrospective causality analysis": figuring out why complex chains of events happened and then trying to do things in the future to profit from that knowledge. As ever, I try to get more examples of what you'd use this kind of tool for. There's also some storage optimization things with Hadoop.
• Beautiful Data book coming out, I have a review copy on the way. Also, the video from John, like this one of Chris Curtin.
• Personal metrics crossed with the old BMC airport example.
• RightScale: doing DB2 management, RightLink (chef plus Nanite).
• Open Source Cloud Computing Forum with RedHat - which explains all of John's KVM, Xen, libvert Twitterings... cobbler & puppet, and more! And yet, John says, there's no uber-cloud strategy from RedHat.
• I ramble on about how the adoption of public cloud computing in big enterprise accounts are not culturally ready for it: it's the "our customers are not asking for it" answer.
• In enterprises, we need some more CTO input in addition to the CIO role: innovation vs. keeping the lights on.
• We do some cross-podcast pimping to The Agile Executive, esp. the recent podcast episode on Agile Operations.
• Spunk 4.0 and the consumerization of IT.
• I was at the Adobe Industry Analyst Summit this week (see here, here, and here for more), where their CIO showed off a custom UI on-top of their service desk. This raises the question: why did she have to go through customizing it? Why aren't service desks good looking already?
• We go over the OpenSourceWorld and CloudWorld conference coming up. I've got a code for free passes if you want one. John's going to a Java conference in Bulgaria to speak to cloud. And also, Antwerp.

Disclosure: IBM is a client, see the RedMonk client list for other clients mentioned.

You can download this episode directly directly and it'll also show up in the RIA Weekly feed for iTunes and other podcatchers. Or, just use the controls below to listen to it right here:

This week, Ryan and I go over the recent RIA news. My notes aren't as in-depth as the usually are, but here's the high-level outline:

• Ryan's mountaineering. iPhone 3GS - how's that video? We review YouTube vs. blip.tv vs. Vemeo.
• Personal data analytics - makes me remember one of the Azure contest winners, http://myimpulselive.com/.
• Silverlight 3.0 launch - all sorts of overviews out there, and also see the RIA Weekly video if you haven't already.
• SketchFlow, and SketchFlow vs. Catalyst.
• Continental Case with Silverlight; see also Cynergy examples along these lines.
• I give a re-cap of the Adobe Industry Analyst Summit I was at this week. See posts here, here, and here.
• Summary of the Matthias Zeller podcast on portals and composite RIA.
• In talking about a story on Mixpanel, Ryan busts out with "Interaction Analytics" and recalls like Robot Replay. We also discuss how lacing in interaction analytics seems like a natural for RIAs.
• How can we get RIAs to improve the software development process? Seems like Agile and developer-centric cloud computing would converge with RIAs nicely.

Disclosure: Microsoft and Adobe are clients.

While at the Adobe Industry Analyst Summit this week, I caught up with Matthias Zeller. Having been at Adobe for sometime, he's had an interesting history with RIAs from the work he did with SAP (where he's a mentor) and now onto the Adobe Genesis project (see blog too) that he's been working on in recent times. As Matthias says, he's been traveling around to talk with customers a lot recently, specifically around how they use and would like to use portals and situation applications.

Ever since talking with James Ward back in episode #41, I've been keeping an eye on how RIAs might could be used for portals. In going over the customer conversations he's been having, Matthias gives us a pretty good idea. He uses the term "composite RIA" several times, which is a nice follow-on from the "mashup" and "situation application" phrasings.

Disclosure: Adobe is a client.

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:

This week, John and I have a fun guest on, Rob England, aka, The IT Skeptic:

• How to tell the difference between Australian and New Zealand accent.
• How did Rob get started doing the IT Skeptic.
• What's up with ITILv3? Do you think we'll ever get ITIL for free? Do things like MOF being Creative Commons put pressure on ITIL to be more open. Rob also suggests COBIT.
• The lean CMDB - having a maximal CMDB costs too much, so you need to have a scale-backed one of some sort.
• What's the gradient of CMDBs out there? The ITIL definition is very clear, but there's lots of people on "the journey" to CMDB. The emphasis for CMDBs is on maintaining the relationships between raw IT and the business services they help deliver.
• How does CMDB auto-discovery mix with reality? An initial baseline/theory is good with discovery, but the problem is what's discovered may not match "reality" - how things are supposed to be - missing all the rouge configuration out there. Process has to be applied to keep things discovered and modeled properly. Ongoing, discovery is good for auditing your assumptions about IT, but maybe not the best way to get the "pure" CMDB model. And then there's all the manual stuff as well, like mapping up the business services.
• What will be impacted by this change? What we need is a configuration process, not a magic tool. The "magic tools," of course help the process, but the process is the overlord. While on the one hand hand, you don't want your CMDB walking out the door - that is, it being all in employee's heads - on the other hand, sometimes that works fine if the process is optimized.
• While we know that change management is good, it seems so painful to do it, so how do you get people to start doing it? "I think change is all stick and no carrot, unfortunately."What's the aspiration vs. usage of ITIL Rob's seeing out there? Lots of people know about it, and aspire to it, but usage... there's 1/2 million people with ITIL foundation training, which means a common language is (probably) being formed.
• John asks about crossing IT Service Management with cloud computing. Esp. of interest here is how a service catalog fits in and linking up cloud stuff with business services. With cloud, the ITSM problem is that you don't have visibility into what's going wrong with your service providers to diagnose problems.
• So what are the other problems with cloud computing cross with existing IT departments? Testing and change management of existing, even legacy IT.
• Having pointed out the problems with cloud computing, we delve into the benefits and try to rig up a sense for the long-term spread of cloud technologies in IT. Rob says it'll be 10 years before the cloud is a mainstream approach.
• The IT Skeptic books: Owning ITIL, Introduction to Real ITSM, The Worst of the IT Skeptic.

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:

John and I caught up earlier in the week. Despite it being a short time between this episode and the last, we found plenty to talk about:

• Man! It's hotinhur!
• John's Cloud Week videos - John Adams from Twitter, Chris Wensel on Cascading, ControlTier.
• The "science" meme at Velocity - R, Mathmatica, and doing stats. Coté also remembers that Sara Dornsife now works for the R company, and the Zed Shaw talk that included R. I ask John what you'd use this stat stuff for, obviously charting historic data, but also doing predicting.
• Chef vs. Puppet - also cfengine is mentioned in this context. What other "alliances" does OpsCode have around Chef? EngineYard, (maybe, if John remembers) RightScale. Also, they got $2.5M in funding (we mistakenly remember $5M).
• Eucalyptus - talking through their layered up, (seemingly) "swappable" cloud platform and how they're thinking of making money off it.
• IBM CloudBurst demo (this one, or this one, or another one? - how was that? John likes the Service Management hook-in, self-service portal.
  
• Is "private cloud" slowing down "cloud computing." And getting at traditional IT shops "hav[ing] that ability to, kind of, 'API' themselves."
• Are SLAs and KPIs calling out for "better" operational stuff in IT like private cloud?
• This gets us into webappVM - deep instrumentation; "webappVM is building a self-monitoring application cloud for web applications, bringing enterprise-level application management capabilities to public and private clouds" from their site.
• FiveRuns Dash - generic-ish, Web 2.0 style metrics. This gets us onto FiveRuns memories. Also, former FiveRuns CEO OT is now at Zenoss as their CMO.
  
• Google Ops Head goes after blood at Structure 09, .Net at MySpace
• Kuleto's review
• John [was] helping run CloudCampColumbus this week. He's trying to get some Tivoli folk out there.
  
• Coté interested in pictures and anecdotes of Detroit. [We'll see if John snatches some.]
• I mention that I'm thinking of strumming up interest in barcampESM II (see the first for reference) - is there enough interest for that kind of thing this year? Also, I'm lazy.

Disclosure: Reductive Labs (Puppet), IBM, and Zenoss are client.