Interesting tidbit from the folks at Geeknet (They are the owners of  SourceForge, Slashdot and Ohloh). OSS projects hosted on SourceForge are increasingly becoming more platform agnostic. Since early 2005, the percentage of projects that are platform agnostic has increased from approximately 50 percent to nearly 70 percent as of Q3 2009.

Here are some of the numbers:

• Amount of OSS Windows compatible projects has grown from 72 percent in 2005 to 82 percent in 2009. (This is approximately 350,000 of the total 433,000 OSS projects)
• Windows is the only operating system compatible with the top 10 all-time most downloaded projects on SourceForge
• 23 of the top 25 projects run on Windows and 14 of them run only on Windows

Related Articles

• Geeknet at OSBC 2010 (port25.technet.com)

Technorati Tags: Geeknet, Microsoft Windows, Open Source, Operating system

Zenoss Core Dashboard

This week I am very excited because we released Zenoss Core 2.5.2 with a cool new feature,  Xen hypervisor monitoring.

I am very happy to see Zenoss put this out as open source software, because it’s the beginning of what could be a great piece of software. Now that this is out there for anyone to use I expect to get lots of feedback to extend and improve it.

This extension to Zenoss Core discovers guests on Xen para-virtualized hosts and provides monitoring of performance and availability via SSH. While the Xen hosts run on physical servers, the virtual guests are listed per host and linked back to any discovered instances on the network. Admins can quickly find the associated hosts and guests and monitor their Xen virtual infrastructure along side their entire physical, virtual, and cloud-based IT environment through a single interface.

I really think this is the beginning of a lot of virtualization and cloud monitoring capabilities from Zenoss Core. In recent weeks we got a couple neat extensions for virtualization and cloud monitoring including Amazon Web Services,  Google App Engine, libvirt (a virtualization library favored by Red Hat) and Ganglia integration.

Technorati Tags: Hypervisor, Monitoring, Xen, Zenoss Core

On February 6, the White House Open Government Initiative launched a government wide public participation opportunity.  As part of the Open Government Directive issued in early December, every major agency published an open government website.

In response the Open Source for America (OSFA) initiative has submitted their:  “Proposed Guidelines for Open Government Plans” to the Obama Administration for consideration. These plans are meant to encourage the government to embrace the open source ideals of participation, collaboration and transparency.

The proposal notes that there is, ” ….a tremendous opportunity to capture the innovation and ingenuity of government employees, who have the means to create their own tools to make themselves more effective, rather than waiting for a cumbersome and unresponsive procurement process. Open source software is, in fact, the most concrete form of participation available to the government’s constituents and its employees”

To get involved the Administration has invited public comments on each agency’s open government page through March 19, 2010. Open source advocates are encouraged to visit these agencies listed on the White House website and share their opinion.

Technorati Tags: government, Open Source, United States

Rube Goldberg would be proud of this machine.

Technorati Tags: cadbury egg, funny

In a time when many tradeshows are experiencing lower then normal attendance the 8th Annual Southern California Linux Expo (SCaLE 8x) had record attendance this past weekend in Los Angeles. I was there exhibiting and conducting a community training day for Zenoss and was very impressed by not only the quality of the program but the enthusiasm of the attendees.

Here are some of the highlights:

The Mini Conferences

On the Friday before the main SCaLE expo and speaking program starts many people hold mini-conferences and the SCaLE staff has been excellent at helping to organize and promote the events. I conducted a  Zenoss Community Day that Friday with phenomenal attendance and enjoyed meeting a great group of open source management users. I also peeked in on Ubucon which had a standing room only crowd of Ubuntu users. There were also quite a few other special events that seemed to be well attended.

The Keynotes

I got to watch both keynotes this year and they were both excellent. Here’s a little recap.

Karsten Wade’s – Being a Catalyst in Communities – The scientific facts about the open source way

Karsten’s a bona fide community builder helping drive the growth of the Fedora project, an open source distribution sponsored by Red Hat. His talk was a great overview of how to drive community participation and better yet, what results not to discount. He also announced the newly published free book, The Open Source Way – Creating and nurturing communities of contributors. A blue print of how to apply open source principles to communities and facilitate participation. In his presentation he made reference to an initiative sponsored by Red Hat,  Professors Open Source Summer Experience (POSSE), to help professors understand how to get their students involved in open source which was very cool.

He also made some interesting references to research done by Etienne Wenger on Communities of  Practice, which are groups of people who share a concern or a passion for something they do and learn how to do it better as they interact regularly. Which is part of the science indicated in the time of the talk.

Tarus Balog’s – So, You Think You Want to Start an Open Source Business

Tarus is the lead of the OpenNMS project which he maintains in conjunction with a services business, The OpenNMS Group. He offers a very candid tale of his starting the OpenNMS group using the board game, Life, as a metaphor. It was very clever. Tarus and I share a common interest in open source IT management given the companies and projects with but we have some different philosophical views on how to develop those companies and communities. Despite that I really respect his passion for his project and his company. I thought his presentation was very well done and really enjoyed his talk.

Both keynotes are available here on UStream.

Sessions

I got to sit in on Stephen Spector’s presentation on Xen and had a lot of time to talk to him through out the show. Xen fascinates me as a open source virtualization technology it’s used everywhere and even serves as the infrastructure for Amazon EC2.

I was there to man the Zenoss booth so I missed out on a couple of scaling and cloud presentations but the word was they were all great:

• Ari Lerner’s presentation on Pool Party. Written in ruby, PoolParty provides a nice domain specific language for describing a repeatable, declarative cloud computing infrastructure. Mainly focusing on amazon’s EC2 offering, the presentation will cover basic concepts of cloud computing, how PoolParty works and how you can get into the clouds in one command
• Scaling Facebook via Open Source – Given their use of  of the following open source projects in highly available deployments I thought this would be interesting: Cassandra, Hive, Haystack, memcached, MySQL, PHP, Scribe, and Thrift.

If you are an open source fan or vendor and can make the trip, I highly recommend attending to SCale 9x next February.

Related Articles

• SCaLE 8x Report – (linuxjournal.com)
• Improving the FLOSS legal landscape (iquaid.org)
• SCALE8X Day 2. Ubuntu Kernel Hardware Compatibility Testing. (canonical.com)
• SCALE8X Final Day Ubuntu Hardware Compatibility Testing. (canonical.com)

Technorati Tags: Fedora, LInuxfest, Los Angeles, Open Source, OpenNMS, SCALE8x, Xen

How do you build one of the busiest websites on the Internet? You wouldn’t guess the right answer to be, “You download some free software and hack  it”…Actually  the question is how do you build one of the world’s busiest websites that will scale affordably? You use open source software.

Twitter showed everyone their cards recently by publishing all the open source projects that they are contributing to. This is the picture of how open source software should work.

Organization has a a big, hard problem to solve. They write some software or update existing software and then publish what they did for others to use.

The people writing the software are writing it to solve a real problem. Not to sell it to people with real problems or to prevent the problems a clever sales guy manufactures. Given Twitter’s massive amount of users and traffic it’s probably one of the most demanding IT environments in the world. If these are the people making the software than you have some idea that it an stand up to scalability issues in other use cases (Failwhale’s not withstanding).

Not only do the Twitter engineers donate their work to the open source community they also share their insights on what happens in their environment on the Twitter Engineering Blog.

As I look at what the Twitter guys are working on it’s pretty impressive and has a fairly substantial impact on high availability computing. Some of the projects that caught my attention:

• scribe.gem – A Ruby client for the Scribe distributed log server.
• contributions to memcached.gem – An interface to the libmemcached C client. (memcache is a distributed systems for improving performance across dynamic web apps)
• contributions to cassandra – A highly scalable second-generation distributed database.
• contributions to hadoop – A platform for analyzing large data sets.
• contributions to pig – A platform for analyzing large data sets.

Any how it’s good to see open source users giving back their improvements, I love it when a plan comes together.

Related Articles

• Twitter Loves Open Source And Launches A Directory To Prove It (techcrunch.com)

The folks from the Cloud Computing Manifesto , a group dedicated to open standards for cloud computing, have released a new whitepaper, Cloud Computing Use Cases White Paper version 3.0.

Here’s their description of the paper:

Version 3 of the White Paper addresses the importance of Security in the Cloud. Security in the Cloud is consistently raising the most questions as consumers look to move their data and applications to the cloud. Cloud computing does not introduce any new security issues; the issues that apply to cloud-based solutions are the same issues that apply to in-house solutions. The discussion of security in the paper covers the security controls and federation patterns that cloud providers must deliver.

The paper provides a good overview of cloud computing and provides a taxonomy that is useful for having a discussion of cloud computing. There’s also a pretty active Google Group for discussing cloud computing use cases and discussion of the paper worth watching as well.

Technorati Tags: cloud computing, Security, use cases, White Paper

I have been watching Amazon EC2 progress but I had never deployed one on my own. I really wanted to see if it lived up to the hype. So far it does. I have been using an Amazon S3 account already to back up my blog and some other files and wanted to see if the EC2 experience was equally painless.

Amazon Elastic Compute Cloud (Amazon EC2) is a web service that provides resizable compute capacity in the cloud. It is designed to make web-scale computing easier for developer but for tinkerers like me it’s a great way to bring up a test or lab server. Since you are charged by your compute hours you only pay for what you use not for taking up rack space in a data center as you would with a hosted server. I expect for my needs I can probably get away with running my test servers for a few bucks a year. I am sure persistent servers would only be slightly more expensive.

For my first test I needed something to deploy. Amazon has a huge directory of Amazon Machine Image or AMIs.  I decided to use one of the cloud images from Btinami,  a Joomla CMS image for a project I was working on.

Setting up the account for me as an existing Amazon customer was a piece-o-cake. I just signed up and then entered my telephone number, Amazon’s automated system called me immediately and asked for a PIN number used during the sign-up process. Next I logged into my EC2 Account and configured my security group – I added support for HTTP (Port 80), SSH (Port 23) and HTTPS (Port 443) since they were the ports I might need to access Joomla! and the server. Then I went through the process. I got so excited about how easy it was I followed up and installed an Ubuntu image for another project. I documented it as I went figuring it would have some level of complexity. However I was wrong it was dead simple. Here are the 1o easy steps for deploying a AMI.

Step 1: Find an Amazon Machine Image

I chose this Ubuntu 9.10 Server 32-bit (Karmic Koala) uploaded by Canonical from the AMI directory and made sure to grab the AMI ID from the listing.

Find an Image and then Remember the Amazon Machine Instance ID

Step 2: Set-Up an EC2 Account

Set-up a Amazon EC2 Account and login.

Amazon EC2 Dashboard with Launch Instance Option

Step 3: Choose your Amazon Machine Image

Choose your Amazon Machine Image (AMI) by clicking on the Community AMI Tab and entering the AMI ID from the listing you choose in Step 1.

Select Your AMI

Step 4: Request You Instance

Next you can choose the number of instances by size and billing options. You have a number of choices to make here including which data center you want to run your instance from (I had four choices). I also could chose between two billing options.

Launch Instances

EC2 Instances let you pay for compute capacity by the hour with no long term commitments. This transforms what are commonly large fixed costs into much smaller variable costs.

Spot Instances

Spot Instances let you pay for compute capacity by the hour at a Spot Price that fluctuates based on supply and demand. You specify a maximum price you are willing to pay per hour, and your instance only runs when the Spot Price is at or below that price. This allows for cost reduction on compute tasks with flexible start and end times.

Request Instances of your AMI

Step 5: Advanced Instance Options

I chose cloud monitoring here, for an extra fee you can have Amazon monitor your EC2 Instance. I checked the box because I was interested in what they offered in the way of monitoring:

Step 6: Create a Key Pair

Create a Keypair

Step 7: Configure Firewall

Configure Firewall

Step 8: Review and Launch

Review Settings

Step 9: Wait for the image to boot.

Success!

Step 10: View Your Amazon Instance

View Instances in EC2 Dashboard

Technorati Tags: Amazon Elastic Compute Cloud, Amazon Machine Image, Amazon S3, cloud computing, howto, Web service

I have been a little bit stubborn on my concession that cloud computing is really going to be the wave of the future. Though lately I have been playing with a lot of cloud related technologies and am hooked. Spin up 10 (or 100) Linux servers in a minute from a web page with no real systems management skills and then you start seeing the power of cloud computing. As I try to get smarter on cloud computing I compiled the following list of resources that I have found helpful.  If you have a great cloud resource please feel free to let me know what I am missing on.

Cloud and Virtualization Guides and Whitepapers

• Ars Technica Guide to Virtualization: Part I – From 2008 but still a decent primer on virtualizaton
• The Ars Technica Guide to I/O Virtualization – From February 2010 very nice thorough intro to I/O virutalizaiton
• A Guide to Getting Started with Cloud Computing - Oracle/Sun has a nice overview of cloud computing (peppered with their products and suggested strategies of course)
• Cloud Computing Use Cases v 3.0 – Good collaborative paper from the folks at the Open Cloud Manifesto

Various Cloud Resources (Podcasts, Mailing Lists, Wikis)

• Cloud computing Podcast – A good podcast on cloud computing
• Federal Cloud Computing Wiki – U.S. government cloud computing resources, informal but interesting
• Cloud Computing Wiki – Not all that organized but some nice nuggets in the links
• IT Management Podcast – All things system management but lost of cloudy content.
• Reservoir – Resources and Services Virtualization without Barriers is a European Union FP7 funded project that will enable massive scale deployment and management of complex IT services across different administrative domains, IT platforms and geographies.
• Cloud Computing Email List: An often lively email list discussing cloud computing.
• Cloud Computing Blogs & Resources. An excellent and big list of cloud resources.
• Cloud Computing Portal: A community edited database for making the vendor selection process easier.
• List of Cloud Platforms, Providers, and Enablers.
• datacenterknowledge.com’s Recap: More than 70 Industry Blogs : A nice set of blog’s for: Data Center, Web Hosting, Content Delivery Network (CDN), Cloud Computing
• High Scalability’s List of Cloud Blogs – It’s really good list of cloud computing resources

Cloud Computing Blogs

• CloudDB.info – Cloud Computing News and Info from a Database Geek!  Written by Lewis Cunningham.
• CloudEnterprise.info – Dmitry Sotnikov news, analysis, and opinions on Cloud Computing and SaaS in the Enterprise.
• CloudSecurity.org: Craig Balding covers -a Fortune 500 banking exec covers -  What is Cloud Computing Security?  What are the risks, threats and countermeasures?  How do we ’secure the Cloud’.
• Cloud Talk:  John M Willis blog on IT management and Cloud technology. Also, check out the Cloud Cafe Podcasts
• dba-Research Blog – The DSA (Distributed Systems Architecture) Research Group at Complutense University of Madrid
• High Scalability – Nice blog and resource for scalable,  high availability computing
• Elastic Vapor – Reuven Cohen’s popular cloud computing blog
• GoGrid Blog: Blog with product and industry news related to Cloud Computing and GoGrid
• IT Management and Cloud Blog: John Willis’ personal comments and podcasts
• Head In The Clouds:  Written by Bert Armijo, SVP at 3tera, includes product info as well as comments on industry events
• SpoutingShite: Ross Cooney, MD of Rozmic. Cloud Computing isn’t about technology; it’s about enabling your technology led business model with lower capital investment
• TodoOnDemand: Blog about SaaS, Cloud Computing, On Demand Software, Business models, etc…
• dev2ops: Blog by DTO solutions  and open source developers of ControlTier cloud and data center automationexperts
• Cloud Musings – Kevin Jackson from DoD contractor NJVC talks about clouds
• News and Ideas about Cloud Computing: Edward M. Goldberg reports about ideas and updates to the Cloud Community.
• ReliaCloud: One of the largest cloud platforms in the United States, on news and updates in the cloud world.
• Randy Bias – Cloud Scaling blog from people who implement cloud infrastructure for a living
• Rightscale Blog – By the people who bringyou the Rightscale Cloud Management Platform
• VMBlog – Excellent source of Virtualizaton and  Cloud Computing News
• VMware Community Blogs – Blogs about virtualization and all things VMware

Twitter

• A list of cloud computing Twitterati

Technorati Tags: cloud computing, directory, Distributed computing, SaaS, Software-as-a-Service, Virtualization

So what do you do on a cold winter’s Saturday in Austin? Well if you are interested in Cloud Computing and IT operations you go to OpsCamp. This past Saturday was the first of what we hope to be many OpsCamps, held in an unconference format, to discuss ideas around next generation technologies and strategies for IT Operations.

The OpsCamp crowd was an eclectic crew of systems administrators, open source software developers, traditional software people, and service providers who came together to figure out how to address the needs of IT operations and the tools they use to administer virtual/cloud systems.

One Theme – Everything Old is New Again

I sat in on quite a few conversations most centered around two old themes: Monitoring and Configuration Management. While the problems are similar the landscape is changing. The use of the term devops was very prevalent noting that systems administrators are often required to have some degree of scripting and software development skills to adequately do their job. Effective systems administration goes beyond keeping servers running but the automation of tedious process to increase productivity. In the monitoring arena concerns centered around the amount of data we collect and how to make that actionable to assure consistent service levels.

Lightning Talks

The event started out with a word from the sponsors who had five minutes to talk about their products and services or otherwise woo the crowd. The main sponsors each took their shot at defining their interests in the cloud and how they could help during five minuted lightning talks. Presenters included Erik Dahl (CTO of Zenoss), Brett Piatt (Technical Alliance Manager at Rackspace Hosting) and Luke Kanies (Founder of Puppet and Reductive Labs).

Thanks to the other sponsors Bitnami, Opscode, Spiceworks and Redmonk who also helped make the event happen.

The Unpanel

Rather than having a meticulously planned agenda OpsCamp is interactive. During the Unpanel members of the audience got on stage and talked about what they thought would be interesting topics to discuss this included OpsCode CTO  Adam Jacob, Zenoss CTO Erik Dahl, Puppet founder, Luke Kanies, Travis Campbell from the University of Texas, Systems Administrator Matt Lawrence, Damon Edwards — President of DTO Solutions, Andrew Schafer and moderator extraordinaire — John M. Willis. This led to an interactive session with the rest of the OpsCamp attendees to address the following questions:

• What is DevOps?
• What is monitoring?
• How do you monitor/manage/apps/systems that use multiple environments
• Should I be listening to events or actively monitoring systems?
• Have you thought about monitoring from the cloud into your data center?
• What are the new things that need to be monitored?

The panel discussion was very interesting as well as the audience inputs. The session runs about 30 minutes but is pretty interesting to get the takes on cloud computing challenges from some exceptionally bright people.

The OpsCamp Session Board

Sessions

After quite a bit of discussion the sessions were decided on the following sessions:

• Agent Evolution – Do we need agents for management and if so are they going to be heavy, tedious and burdensome as the agents we have come to know and love.
• The Evolving Role of the Sysadmin
• Monitoring for the Cloud or Service Level Assurance
• Toolchains for Clouds
• Vendor “schmoozing” to talk about the products the vendors had to offer for cloud computing

The Organizers

Even though the conference was an unconference I think a big nod needs to go to my fellow organizers (or my co-conspirators) who worked together virtually from all over the country to pull of the event. Many thanks go to Damon Edwards (DTO Solutions and part of the Control Tier project),  John M. Willis (cloud guru from OpsCode) and Dave Nielsen (Founder of CloudCamp),

Videos and Pics

Damon has some videos from the day with Luke Kanies of Puppet and Reductive Labs and Bill Karpovich, CEO of Zenoss on the Dev2ops blog. We also have a bunch of pictures from the event in the OpsCamp Flickr group.I created an OpsCamp Twitter list for those of you looking for some folks who are smart on cloud computing and devops.

Keep your eye on OpsCamp.com for future events. I believe this one was one of many to be held around the world.

Related Articles

• Opscamp Austin Roundup (johnmwillis.com)
• OpsCamp Austin: IT Ops and the Cloud (cloudswitch.com)
• Videos: Luke Kanies, Bill Karpovich, Ernest Mueller at OpsCamp Austin 2010 (dev2ops.com)
• OpsCamp Debriefing IT Management & Cloud Podcast #66 (redmonk.com)
• Something about Austin #opscamp (austin.metblogs.com)

Technorati Tags: cloud computing, CloudCamp, Configuration Management, Open Source, Opscode, Reductive Labs, Zenoss