ABLEconf Free Software for Free Enterprise

A content management system is a centralized repository for your content. A Web site is a composite of decentralized fragments that are assembled on the edge, in just-in-time fashion as the content is being delivered to users. If it's not a Web site, what does a CMS do? more>>

Creating a live Linux USB stick isn't anything new. And, in fact, the ability to have persistence with a live CD/USB stick isn't terribly new. What many people might not be aware of, however, is just how easy it is to make a bootable USB stick that you can use like a regular Linux install. more>>

Razor-qt is a new desktop environment based on the QT toolkit. I installed it from the PPA and gave it a quick go. It’s early days for the project, but it might eventually become a refuge for lovers of KDE 3 in the same way that Xfce has become popular with people who want to recreate the Gnome 2.x experience. more>>

SCaLE 10x was a grand time. Thanks to all of the volunteers that make it happen every year for us! SCaLE adds new things and puts on a very organized, successful experience every year. The volunteers who make it happen are amazing.

read more

Portability is a huge advantage that Linux enjoys over most other desktop operating systems as most major distros work very well when installed to a flash drive. However, there are still machines in service that just won't play ball when it comes to USB booting. Fortunately, I found a nice little utility that can work around this problem. more>>

Perl has been around for more than 20 years. During that time, it has received its share of both praise and criticism, and lots of misconceptions surround it. Much of this stems from long-outdated notions of what Perl used to be, but have nothing to do with what Perl actually is today. more>>

I've realized I've missed out on a huge area of computational science—chemistry. Many packages exist for doing chemistry on your desktop. This article looks at a general tool called avogadro. It can do computations of energy and gradient values. Additionally, it can do analysis of molecular systems, interface to GAMESS and import and export from and to several file formats. more>>

SCALE team member Gerald Fontejon met with Chris Nolan (eHarmony) to discuss DevOps and the nVentory project. Chris Nolan and his colleagues will be presenting on nVentory at DevOps Day Los Angeles on Friday January 20th.

SCALE:From looking at who is presenting the topic "nVentory - Your Infrastructure's Source of Truth", we have a Director, System Administrator, and Developer presenting - how does each role play a part with the implementation, management, and daily-use of nVentory?

Chris: I initially implemented nVentory shortly I after I started at eHarmony 18 months ago. There was no easy way to track all of our servers and it was frustrating coming in new and not knowing what was in our datacenters. Ongoing I have become more of an evangelist of nVentory, both internally to eHarmony and externally to the internet community.

Darren, as the engineer, works on features and bugs for nVentory. He is an active developer and sends a majority of his work back to the community, allowing others to share in our new features and tools.

Jeff, as a SysAdmin, is a big consumer and user of the information stored in nVentory.

SCALE: Also could you describe your environment at eHarmony in which you guys use nVentory?

Chris: Our environment is a mix of physical and virtual machines, or nodes, with hundreds of applications and data stores. We use it to manager our thousands of nodes (both physical and virtual) and layer metadata on top of the nodes and nodegroups to manage apps, monitoring, release, etc.

SCALE: It does give you some flexibility with the tool having Darren Dao adding new features to fit your team's needs.

Chris: Definitely. Having the developer of a tool in house is great; especially someone as gifted as Darren.

SCALE: Are the other tools that are comparable with nVentory?

Chris: For tools that do just what nVentory does there are not a whole lot of open source options out there. Puppet and Chef have a "server" concept that allows you to query information about your hosts, but neither track - storage, VIPs, network ports, etc. like nVentory does. There are enterprise solutions but nobody likes shelling out wheelbarrows of cash and working with stuffy consultants. :)

SCALE: Would you be able to use another tool like openLDAP to be the "Source of Truth" within your environment by injecting/adding new fields to accommodate your informational needs?

Chris: LDAP doesn't offer an API or an elegant CLI to extract information. It also doesn't have all the fields or an easy way to layer metadata on top of the nodes it may be tracking.

SCALE: I noticed from poking around LinkedIn, all the members possibly met at ATT Interactive, and had some experience with nVentory in the ATT environment. From your experience as a team of implementing nVentory within eHarmony, what were the major challenges?

Chris: Jeff, Darren and I actually met at a startup that did out of home advertising on digital billboards and TV screens.

Though I wouldn't call it a major challenge, the long haul has been getting people to use it. Sometimes people have to ask a few times about what node is handling service X or what port node Z is plugged into. When you keep telling folks to use nVentory and get the data there, they eventually get it and are more productive in the end. It also results in fewer shoulder taps for the Sysadmins, which is always a good thing.

SCALE: Agreed. Always good to provide users the necessary tools to find information on the given infrastructure. And having the information accessible by everyone outside the SysAdmin team is awesome. Researching the benefits of a "Source of Truth" concept, and nVentory - I'm thinking the number of its implementations will grow. However not many companies have thousands of nodes in their datacenter. What number of nodes would you say is the threshold to make a tech dept. invest some man hours on its 'source of truth' utility implementation? And given your experience with nVentory, what would be some of the crucial steps in order to implement nVentory. The datacenter environment touches multiple teams (netops, sysadmins, security, appops), so I would assume there has to be a lot of coordination.

Chris: I think this question can be asked of all sorts of "devops" tools. When is the best time to implement configuration management? When should I start using Jenkins or a similar Continuous Integration tool? When should I start doing test driven development? And finally, how large should my infrastructure be before I start using nVentory?

If you are only managing 1-10 hosts and don't expect to scale that then you probably don't have any problems that devops (and nVentory) address. If you have 10+ hosts and you expect to scale that as you grow then you should implement nVentory now (as well as the surrounding tools and practices I mention above).

Coordination between groups is important. For example, a lot of our tools make decisions based on rack location. This, unfortunately, is not automatable and requires that DCOps be onboard to manually enter the information. When people see the benefits of using nVentory they are happy to oblige and do what is necessary to make the data useful and valuable to the company.

SCALE: What do you mean by, "a lot of our tools make decisions based on rack location?" I'm unclear on this. And also a couple of the devops tools you mentioned, such as CM and CI are cloud ready - meaning these tools could work within a company's an EC2 or RackSpace environment. Is nVentory capable of supporting a hybrid physical/virtual datacenter(s)?

Chris: Let me clarify that statement. A lot of our tools use rack location in their configuration and deployment decision making.

One example is our custom private cloud. When a multi-node requests (a single service needing multiple services for load balancing and redundancy) the system decides where to place the VMs to ensure they are on different hypervisors in different racks.

nVentory is capable of supporting hybrid environments. The considerations would be around networking and encrypted tunnels between sites and providers.

SCALE: Well, thank you Chris for the interview and a glimpse of eHarmony's utilization of nVentory. I look forward to checking out your discussion. For anyone that is planning on attending your team's presentation - you have homework to prepare them for your guys' talk? Or will the discussion be relatively self-explanatory?

Chris: It's been fun, Gerald. People should think about the foundation tools they use to manage their infrastructure and come ready with questions. Our talk will be self-explanatory with examples of how we integrate our tools and infrastructure management together using nVentory.

[Blogger's Note: SCALE 10X will be asking some of the speakers to weigh in on their presentations for the expo. This is the Q-and-A for Mark Burgess, who will give a talk on “3 Myths and 3 Challenges to Bring System Administration out of the Dark Ages” on Friday.]

Q: Could you please introduce yourself and tell us a little about your background?

A: Well, let's see. I'm the founder, CTO and original author of CFEngine. I also used to be Professor of Network and System Administration at Oslo University College, but now I'm full time at the CFEngine company. I have a Ph.D in Theoretical Physics and I've been working on system administration as a researcher and sort of activist for nearly 20 years. I've written a number of text books for universities and dozens of scientific papers on the subject and I'm still pretty engaged in thinking about the future of IT technology. I like to get my hands dirty with practical issues, even though I'm a theorist. Theory is no good unless it is an accurate model of reality.

Q: You're giving a talk at SCALE 10X on “3 Myths and 3 Challenges to Bring System Administration out of the Dark Ages.” Without tipping your hand on the actual talk, can you give us an idea of what we might expect?

A: The talk is really about the future of system administration or IT management. I suppose what I'm doing is pointing out that system administration need to move on from some old-fashioned practices and explain why. Although it is not directly about CFEngine, a lot of these ideas have really come out of the years of research and lessons learned from making CFEngine 3. I'll be saying why we need to start looking at this kind of third generation technology to handle what's coming in the next ten years. I hope the talk will be a bit entertaining and will poke fun at the mirror in a humourous way.

Q: According to your bio, you have a very extensive background on System Administration, in both the academic and private sector. How has System Administration changed over the years, and are the challenges different than they were in years past?

A: That's quite a difficult question to answer. I feel that system administration itself has not changed very much in spirit. We cling to a lot of old ideas in a majority of institutions. The tools and technologies have changed a lot though -- and have especially been enhanced by Free/Open Source Software. The GNU project was a kind of Berlin wall moment for software -- opening up the borders of software creativity from the old monolithic software vendors to a freedom of ideas and the ultimately to a new commerce. That's now reaching maturity. This was always to be expected, but it's a subversive trend. That freedom drives commercial growth, which in turn drives demand for IT, and the size of installations -- so obviously datacentres are hundreds of times bigger than they were in the past. Scale and complexity are growing all the time and will soon become unmanageable without an attention to proper modelling. People are being forced to adopt modern methods, sometimes kicking and screaming.

Q: So what is the future system administrator?

A: System administration is engineering, even today, So the sysadmin of the future will be a fully fledged infrastructure engineer, equipped with proper tools and analytical skills. He or she will watch over an infrastructure that is more smoothly hidden from view. I don't like this expression "infrastructure is code" used by some DevOps commentators. I believe infrastructure should be more akin to documentation than code -- i.e. something that's there to be used without having to perform surgery on it to make a change. I think that is a difference between CFEngine 3 and other tools for configuration management, for instance, at least in edge cases. Future engineers will also have to deal with far greater numbers of devices at every scale from mobile phones to small embedded devices, up to mainframes and cloud installations. There might be less diversity in operating system, but a great focus on the application layer -- so ultimately greater complexity. I believe we're moving increasingly into what Alvin Toffler called the Third Wave of society -- away from the industrial era and all its monolithic mass production methods, towards a society driven by individual freedoms and diversity. Hallelujah! (laughs) That makes for an exciting challenge for the IT infrastructure engineer. In particular, it means that knowledge management will be a core competence, and there will be a more optimal division of labour between Man and Machine. We're also going to have to reconsider what we mean by security.

Q: Is this your first visit to SCALE? If so, what are your expectations? If not, can you give us your impressions of the event?

A: I attended SCALE 9x last year and gave a talk on the future of configuration management. SCALE is a pretty friendly event, with smart, likeable people. I'm looking forward to coming back this year.

Q: Is there anything else you'd like to add?

A. Apart from small and preferably prime numbers, no thank you, (laughs). I'll just wish everyone an amazing event and look forward to participating.

Thank you for taking part in this brief SCALE 10X interview.

No, it's sadly not a day for shopping. Today, some of the most visited websites are dark to raise awareness of two bills now making their way through the U.S. Congress. more>>

Actually, it's not so much that the Wikipedia blackout is easily circumvented as that it can take effort to see it. To see the blackout required allowing 3 ( maybe 4 ) domains to provide javascript. No javascript gives a message about the blackout and about SOPA/PIPA and the normal english site full of functional articles and links.

read more

At SCALE, in spite of having a lot of content for developers and sysadmins, we've always had a soft spot in our collective hearts for the newbie.   Sadly, we've not been able to muster a lot of material oriented towards the raw beginner in Linux and Open Source. 

Not until the 10th SCALE, that is.  This weekend SCALE will hold its first ever Linux Beginner's Training class.   It's a two-day 12-hour course on how to install and administer a basic Linux workstation.   The students will learn how to install Linux, configure printing and networking, how to install and update packages, and more.  They'll also get an introduction to the Open Source community: what it is, how it works, and how they've become a member, without having to sign up or pay fees!   The instructors, trainers and proctors are volunteers from the IT industry and the Open Source community.

We allocated a large room for the class, enough for 40 students.    However, apparently there's quite a demand for Linux introductory training, as the 40 slots filled up almost instantly, and we have, at this time, almost 20 on a waiting list.

We at SCALE are contemplating a follow-on class to cover the waiting list, but if you run a training outfit in SoCal, here's an opportunity for you: there's a demand for basic, entry-level Linux and Open Source education.  If you offer it, apparently they will come in droves.

While the beginner's class is full, we strongly encourage those new to Linux to attend the Expo.   There are a number of sessions that will help expose you to the Open Source software culture and help you "talk the talk", and the Expo floor has one hundred booths stuffed with people who would love to chat with you about Linux!   (See the list of exhibitors for details.)

Hope to see you at SCALE!

I'm in midtown Manhattan, connected to the Net over my hotel's slow but costly wi-fi connection. Normally when I'm traveling — at least here in the U.S. — I avoid lame hotel connections by using AT&T's cellular data system, usually through my iPhone's "personal hotspot." more>>

BeOS was a much loved and highly advanced desktop operating system that ceased active development in 2001. ZevenOS is a Ubuntu 11.10 based system (with a bit of help from Xubuntu) that attempts to recapture some of the BeOS look and feel. more>>

SCaLE has once again accepted one of my talk proposals. This year they've tagged me to talk about Nginx.

I'm really looking forward to SCaLE 10x. It's a fun, informative conference. They also add something new every year. For instance, this year one of the new things is a youth-driven mini-conference. Along with the annual educational mini-conference, I think SCaLE's definitely a great resource for parents and educators. I certainly hope the kids enjoy the youth conference!

read more