Red Hat People Blogs

May 03, 2008 11:13 PM

Preparar Firefox para un kiosco

userChrome.css Firefox tiene en la carpeta del usuario, y dentro del perfil una subcarpeta llamada chrome y en ella el fichero userChrome.css
Éste permite ajustar pequeños cambios en el diseño como mostrar u ocultar menús, botones, etc.
Por ejemplo, podemos esconder el menú de ayuda con: #helpMenu display: none !important;
Compactando el entorno Yo utilizo una visión más compacta reduciendo las barras, poniendo la barra de direcciones y búsqueda en la de menús: /* Remove the Edit and Help menus (...) - Software Libre / Software Libre, Linux , Seguridad , Documentación

May 03, 2008 01:22 PM

Five things in music that I hate

Happy Saturday! Today, I will list five things in music that I hate. They are not the only five things, and I don't hate them so much as to do anything more than write about them in my LiveJournal. So, if I list something that you absolutely adore, please, don't sweat it. I won't ask you to turn it off if you listen to it in front of me, just know that it isn't my thing.

1. SKA. Seriously, who thought this was a good idea? Its like a drunken horn section hooked up with Billy Idol. Thankfully, this seems to have died, and No Doubt realized that SKA sucked and stopped playing it. Music historians will look back on "Clueless" and be, well, you know.

2. New Radiohead. OK, so I'll admit to liking some of the tracks on OK Computer. After that, however, I think they just gave up. My best guess is that they realized that they had enough fan base who would buy _anything_ they put out, so they just decided to try to see how far they could push it. Their latest single sounds like they've all suffered brain damage and are just beating on their instruments and/or moaning. Don't believe me? Listen to it again, and whenever the vocalist stops grunting, yell out "TIMMAY!". See? It makes the song better (but still, not good).

3. Jack Johnson. His music is the sort of slow repetitive crap that will be playing in hospital waiting rooms for the rest of time. It is totally unoffensive and wholly without substance. The first time I heard one of his songs, I thought someone had put in one of those "Baby" albums where they use small words and sing softly to children, but then I realized that even babies would be bored to death by it. I hope and pray to the elder gods that he does not start a "mellow" trend in music.

4. Reggae music. When this comes on the radio (usually Bob Marley), I usually mock it, then change the radio station within the first forty-five seconds. Perhaps the fact that I'm not a stoner (nor have I ever been) contributes to this, as I'm told that "it is sooo much better when you're high, man". Maybe that explains Jack Johnson too.

5. The Dropkick Murphys. This one is going to cause the locals to arrive at my door with torches and pitchforks, but I really just don't like these guys. OK, you're drunk, angry, Irish, and you really really want to give the local area a shoutout. IN EVERY SONG. The result? Every song sounds exactly the same. In fact, until recently, I thought they only had one song, titled "DRINKING IN BOSTON WHICH IS F***ING AWESOME!!!!". 1970s Ireland is calling, they want your schtick back. I'm told its a Boston thing, which as I've previously alluded to, I don't get.

OK, now you've seen five things in music that I hate. Feel free to comment on them, or, if you'd like, write down five things in music that you hate. I promise, I won't get mad if you disagree, or hate on one of my favorite bands. Except the Foo Fighters, because everyone agrees that they are rock gods.

May 02, 2008 06:05 PM

Linux Fest Northwest 2008

Ah, the smell of the salty air; the crystal clear sky over a beautiful skyline; the busses quietly humming by on their electric powered motors; Starbucks on just about every corner; Seattle, I have arrived.

I love this city, I really do. The people, the scenes, the ocean, the sushi, I just can't get enough of it. I spent a couple days getting back together with my friends from the area; developers, system builders, IT admins, etc... We talk shop, family, weather, and take in a Mariners game at the wonderful Safeco Field. I'm reluctant to leave, but trek north we must for Linux Fest Northwest.

In Bellingham we arrive, somewhat late at night. Driving through the downtown area we spot a large banner hanging across the street advertising the Fest. Times have certainly changed. It's certainly fun to see the influx of geeks mesh with the biker bars and the college crowd. At the hotel you can tell it's fest time. Lobby filled with geeks: laptops, ham radios, smarmy t-shirts abound; excited conversations about kernels and desktops, and rpms, and debs, and who's going to win Alpha Geek this year. Snickering comments about whether or not the hotel wireless will withstand the abuse a hotel full of Linux geeks can throw at it, and a bemused rueful grin is the only answer one gets from the hotel staff (turns out that the hotel internet is pretty unusable by the time we arrive, but there is open wireless somewhere near that still works!).

A few hours later, after having scrambled to finish up my presentation slides, I collapse into sleep. Very excited about the next morning, dreading the early wake up time though. Of course I wake up an hour before the alarm is set, but that gives me time to write this down. Shower, quick hotel breakfast, a few last minute discussions with fellow Festies and we're off to the show!

Day 1

Lots of excited people filing into the college. Booths mostly setup, last minute scrambles and fixes that always happen. Almost double the exhibitor space this year, a welcome surprise. The usual suspects are in place, Novell, Google, Ubuntu, hey look, a Mandriva! Many booths are using Fedora on their display systems, hooray!

I attend a talk from a friend who is now a Google employee. This is a talk about network protocols and the various evil things folks can do along those protocols. Many people seemed surprised at how many vectors there are to attack and will certainly be thinking twice about what protocols they use.

From there I go to another talk, this one on Grease Monkey. Fascinating stuff, however it really showed that one would have to know javascript pretty well to take full advantage of it.

Next up is my first talk, A sneak peek at Fedora 9. Ustream has a video, and slides are here. This talk went great. I got to showcase a lot of really cool features in Fedora 9, as well as give out USB keys pre-loaded with almost Fedora 9. As soon as the audience caught on to the fact that asking questions got you a USB key I started getting a lot more questions. Not just your typical mp3 questions (which in fact wasn't asked at all!) but more interesting ones, like what is IPA and what about the Certificate Server and things like that. The audience was nice and engaged and seemed reluctant for the talk to end which is always a good sign.

I get dragged into the "Alpha Geek" competition where I do horribly, no surprise there. It was still incredibly fun especially since there were a few Fedora questions which seemed a little unfair given me being a competitor.

The after party was as after parties are, but with a twist. This one had a fair number of ways to damage one's self. Brews, pizza, loud music, oh and lots of electricity generating devices. What fun! The party was set in American Museum of Radio and Electricity. Lots of fantastic old devices to look at and play with plus lots of open source consumers/creators with loosened tongues and stories to tell.

After a few hours, the crew I hang out with migrated to our traditional haunt during the fest. A Greek restaurant/bar that just happens to be within stumbling distance of our hotel. Alas they ran out of Grey Goose rather early. Rehashing of the days events, reminisces of years gone by, and lots of jokes and innuendos thrown about. A few friends stop by to share a drink and a story and before you know it the bar is closing and it's time to capture some sleep before the next day begins.

Day 2

A quick blast through Starbucks and we're on our way to the final day of LFNW. I've got a talk in the first time slot, "Fedora Distribution Toolbox" (Video, Slides). Here I get to show off all the cool tools and services that Fedora has created, fostered, adopted, etc.. over the past few years to make our distribution happen completely on opensource software. There were fewer attendees for this talk, partly because they were aware that the talk would be available as a video later where some other talks wouldn't (also, they knew I had no more USB keys to give away). I still got a good number of questions, particularly around koji, transifex, and MirrorManager. A number of people after the talk informed me that they were much more inclined to bring their software into Fedora and work with Fedora as a distribution mechanism after hearing about all the tools we create and the ideals we hold dear. There were a number of other people who were interested in taking our tools and using them for their own needs. I think I'll give this talk more often as it seems to be an enlightening one.

I didn't make it to any further talks that day, I was too busy chatting 1 on 1 with various attendees who had further questions about Fedora or about some software they were writing and would like to get into Fedora. Before I knew it time had come for me to run the Fedora Birds of a Feather. I saw a lot of familiar faces, those that had been in my previous two talks, but a number of new faces as well. The BOF started off a little slow with people a little shy about asking questions but the conversations did finally get rolling. After the 3rd or so question related to my previous talks somebody had the bright idea to quickly run through the presentations again for those that might have missed it. Lightening round begin! I blew through the presentations pretty quickly and still got a few questions along the way. Once again we were way over time but people seemed reluctant to leave. Finally we had to usher them out as the campus was closing and my ride was leaving.

Exhausted I fell into the back of my friend's car. Satisfied with yet another great LFNW trip and sad that I was leaving the area in a few days and leaving my friends behind. This show has always been one of my favorites. It's very "grass roots"; free admission, word of mouth advertising for the most part, planned in open meetings, and very welcoming of speakers and exhibitors. I will continue to go for as long as the show exists and I can get funding to make the trek. The timing of the show makes it a great target to show off Fedora's next release and to talk about whatever seemed important for that release cycle. If anybody lives in the greater northwest area of the US I would highly recommend making the trip to see this show.

May 02, 2008 03:11 AM

Setbacks

The last couple weeks uncovered a few problems in the incremental compiler.

First, suppose you compile a program with the incremental compiler, then recompile it. You would expect to get the same warnings as well. But — whoops — I never thought about this until a week or two ago.

I hate that awful moment of realization. It reminds me of getting in trouble as a kid. “Oh shit”, I think. “What am I going to do? Does this sink the project?”

In this case, there are some options. If the set of warning flags does not change between compilations, I think I can modify GCC to store the warnings with their corresponding declarations. This is a bit of a pain, but nothing too awful — and I think I can avoid imposing a cost on the non-warning case by representing the warnings as tree objects and storing them in the hunk with the other declarations.

If the user does change the warning flags, then what? Record it and recompile, I guess. A similar idea applies to options that change the ABI — because ABI decisions get baked into the tree objects we create, if the ABI changes, we cannot reuse the trees.

My other uh-oh moment has to do with inlining. I got bored by the tedious sub-projects I was working on — integrating pragmas (by the way. If you design a language, don’t design pragmas. Thanks) into the dependency computation, fixing the remaining test suite failures — so I decided today to start looking at incremental code generation. Something fun!

I tried out a quick implementation. If a function is parsed, we arrange to compile it; if it is not parsed, we don’t bother. This won’t work on real programs, of course, since those “missing” functions have to come from somewhere, but this should give a good idea of the possible speedup.

After testing on my typical small test program (zenity), I noticed something odd, namely that recompilations were not as blazingly fast as I thought they should be. (I first estimated the absolute lower bound as the time it takes to preprocess the source files.)

Hmm. A mystery. But first, a brief aside about tools. The compile server forks and runs code generation in the subprocess. I wanted to debug this fork. So, Plan A: use gdb and set follow-fork to child. But… that fails because, although my program does not use threads, it still links in the thread library (relic of my failed threading experiment), and gdb does not seem to handle this well. So, Plan B: maybe ftrace from frysk can help me — all I want to do is see a stack trace at a particular function call, perfect for ftrace. But, the ftrace I have aborts at startup. So I update and rebuild — but there is a build error. I suppose I could have gone with Plan C: stick in a sleep() call and attach, just like I did 15 years ago. Instead I picked Plan D: printf. Not quite as good, since I still need some of that information. Somehow I didn’t feel like Plan E: rip out the threading code and start over at Plan A.

Right now I’m doing a lot of debugging and pretty much every week has a vignette like that. I didn’t do that python stuff in gdb purely for fun.

Anyway. What is going on in the compile server?

What I found is that the code generation process still does some processing on every function, even functions that we intend to drop. In particular it is lowering each function to GIMPLE. I think what is going on here is that GCC is lowering functions and running local optimizations on them so that they can be considered as candidates for inlining. At least, that’s my working theory until I get back to Plan C and dig around a bit.

I’m not totally sure yet what to do about this. I think I will have to go back and rip out the decl re-smashing pass I wrote a while back, and instead find a way to perform gimplification in the server. That way, the compile server can keep the gimplified form for use by the back end. Other than the work involved, and some tricky details in lowering without smashing, I think this will work.

This isn’t going to be pretty, but at least it isn’t a total disaster. I’d like to think this isn’t totally an accident. GCC has undergone a lot of changes in the last five years to make it more flexible internally, and I’ve pushed a little bit more in that direction on the branch. This makes it a bit simpler to change the point at which we put a fork in the pipeline.

It feels a bit strange to write about the mistakes I make. On the plus side, I know how to fix these problems; writing about really unknown problems would, of course, be beyond the pale.

May 02, 2008 12:54 AM

Labeled NFS Requirements Draft Submitted

Dave Quigley has just submitted an Internet Draft to the IETF outlining the requirements for Labeled NFS:

MAC Security Label Requirements for NFSv4 (link)

Abstract

This Internet-Draft outlines high-level requirements for the
integration of flexible Mandatory Access Control (MAC) functionality
into NFSv4.1 . It describes the level of protections that should be
provided over protocol components and the basic structure of the
proposed system. It also gives a brief explanation of what kinds of
protections MAC systems offer and why existing NFSv4 protection
mechanisms are not sufficient.

This draft is a generalization the original Security Enhanced NFS document posted last year, addressing the general need for mandatory access control support in NFS.

NFSv4 currently supports two access control schemes: standard DAC and ACLs. MAC labeling support is required for technologies such as SELinux and OpenSolaris FMAC.

Essentially what's needed is a way to convey MAC labels over the wire (for both setting and retrieving their values), and to be able to enforce security policy using those labels. The server needs to be able to determine the security label of the remote client process when enforcing policy, and all systems need to be able to ensure they understand each other's labels, or be able to translate them. A "Domain of Interpretation" (DOI) attribute is used to determine the meaning of labels, a term which may be familiar to those who've braved the IPsec specifications. The confidentiality and integrity of these security attributes must be protected in transit, while all parties need to be authenticated. We also need to be able to handle the case where either the client or server does not have MAC enabled, and to ensure non-breakage with existing implementations. There's a lot more in the details, but that's the gist of it.

It may seem at first glance that NFSv4 named attributes (NAs) would provide the required labeling functionality, but they're not a good fit. NAs are specifed as opaque to the system and user-managed, while MAC security labels are managed by the system. NAs also do not provide necessary semantics such as conveying client security attributes or negotiation of DOI. There are also issues with attribute namespaces (which are user-managed and unspecified) and labeling atomicity. Another possible approach is to implement Linux/BSD-style extended attributes (EAs), which are simple text string attributes associated with files, in contrast with the NA "subfile" scheme. This would potentially only solve the attribute namespace issue, and is also not a good general solution. EAs are also not currently part of the NFSv4 specification, and it seems like a contentious area in any case.

The current Labeled NFS prototype code utilizes NFSv4 recommended attributes (RAs), which are fully extensible, already exist, and are already used for similar management of metadata (e.g. ACLs). This seems to be the simplest and most straightforward approach.

Once there's consensus on the requirements, the next step will be to develop a protocol specification and hopefully have it incorporated into NFSv4. v4.1 is currently in "last call", so the next candidate would be v4.2, it seems. The prototype code for Linux/SELinux will continue to be developed alongside the standards process.

For those interested in following or contributing to the project, there are several relevant mailing lists:

• Labeled NFS
• IETF NFSv4
• Linux NFS

Dave is hoping to have further discussion IETF 72 in July, and will be presenting on the state of the project at the SELinux Developer Summit ahead of that.

May 01, 2008 06:37 PM

a very bizarre lunchtime.

I went to lunch at one of the towns 49 all-quite-the-samey sushi restaurants that I'd not tried yet. I was seated next to someone who seemed to be some kind of dial-a-psychic. She had tarot cards all over the table, and appeared to be giving a 'reading' over the phone whilst she gobbled down maki. She had a companion who was also on the phone during the entire meal. I don't think they spoke a word to each other. I'm not sure if she was also a psychic. The whole experience was very surreal.

Then they left, and five minutes later the 'psychic' returned looking for her lost credit card or something. Priceless.

May 01, 2008 05:11 PM

How this for Technical Leadership

I think a lot applause has to be directed at Adobe for making SWF and FLV fully published and open to anyone to build upon. This is what I call a fully spec'ed open standard. Not the half-spec'ed ooxml.

I guess leadership in this space clearly belongs to Adobe. Now to get gnash fully functional - or better yet, Adobe open sourcing their flash players and putting it out on a GPLv2 or even a GPLv3 license.

May 01, 2008 04:15 PM

Example of leadership and accountability

I was referred to this article from the MSM which reports on a contractor whose worker fell and died from his injuries was fined by $150,000 under the Workplace Safety and Health Act.

The contractor, a Mr Lam Teck Foo, clearly did not have a wongkanseng moment.

Just contemplating the similarities and how responsibilities and accountabilities are so diverse is quite disappointing.

May 01, 2008 12:12 PM

Where’s Red Hat This May?

North America

From May 6-9, Red Hat will be in San Francisco, CA at the JavaOne conference in booth #734. Here, you can meet JBoss core developers and attend over 15 different presentations and live demos that will show you how to use JBoss technology to solve real-world problems.

Make plans to join us at our Enterprise Acceleration events around North America too. These events provide the opportunity to hear from Red Hat executives about how IT can enable you to transform your business. You’ll also learn how to move beyond legacy, client server and first-generation Java technology and how to do more with less. JBoss customers will also tell you about the success they’ve had in developing JBoss Enterprise Middleware.

Red Hat executives will also be delivering keynotes at several stops on the CIO Forum and Executive IT Summit this spring. During keynotes, our executives will discuss driving innovation and value across the enterprise with open source, and will draw on the needs of today’s CIOs to map the adoption of open source in the enterprise. Hear how Red Hat is helping customers realize better performance from development to deployment.
We’ll be at the following cities:

• May 1, 2008, Philadelphia
  Keynote speaker: Iain Gray, Vice President, Global Support Services
• May 8, 2008, Chicago
  Keynote speaker: Tom Rabon, Executive Vice President, Corporate Affairs
• May 15, 2008, Hartford
  Keynote speaker: Katrinka McCallum, Vice President, Management Solutions Business Unit
• May, 28, 2008, Charlotte
  Keynote speaker: Lee Congdon, CIO

Other North American events:

• May 1, 2008 JBoss Enterprise Acceleration Irvine, CA
• May 6, 2008 Virtualization: Getting from Pilot to Production Houston, TX
• May 6-9, 2008 JavaOne San Francisco, CA
• May 11-16, 2008 VA Infosec Las Vegas, NV
• May 12, 2008 JBoss Enterprise Acceleration Toronto, Ontario
• May 13, 2008 JBoss Enterprise Acceleration Montreal, Ontario
• May 14, 2008 JBoss Enterprise Acceleration Reston, VA
• May 15, 2008 North Carolina Research Campus keynote by Tom Rabon, Executive Vice President, Corporate Affairs, Kannapolis, NC

See more Red Hat North American events here.

EMEA

• May 6, 2008 JBoss Tech Day, Rome, Italy
• May 13, 2008 JBoss Tech Day, Milan, Italy
• May 14, 2008 Open Source Forum 2008, London, UK
• May 21, 2008 JBoss Welcome, Wien, Austria
• May 28-31, 2008 LinuxTag, Berlin, Germany

See more on Red Hat EMEA events here.

APAC

For information on Red Hat’s APAC events here.

Latin America

For information on Red Hat’s events in Latin America, visit here.

May 01, 2008 12:09 PM

Australian Open Source Industry and Community Report Points to Bright Future for Open Source in Australia

Open source now has an established presence in the Australian market and shows significant opportunity for further growth. As recognized in the recently published Australian Open Source Industry and Community Report, when it comes to open source, the continent is also a key center of innovation and world leadership.

The report was compiled by Waugh Partners based on research carried out in late 2007 and it recognizes the many ways that Australia has made its mark on the numerous facets of open source – industry, community, education and skills, careers and employment, innovation, the market and business development.

Some key points in the study:

• Open source is ‘a key platform for Australian innovation and productivity’
• A number of Australian open source companies are finding success in local and export markets
• International organizations are acquiring open source experts from Australia
• Local contributors are rated among the leaders of the global open source community
• The report also estimates the value of the open source industry to be around AUD$500 million (revenue per annum), with over 50 percent of that directly related to open source.

With new opportunities for development and innovation, the future for open source in Australia looks bright. The Australian Open Source Industry and Community Report confirms a strong market today and forecasts encouraging growth tomorrow.

Working with over 800 customers and over 100 partners in Australia, Red Hat is committed to working with the industry to foster this growth in Australia.

The full Australian Open Source Industry and Community Report is available here.

May 01, 2008 10:34 AM

OOo gio integration

So I finished off Jan’s initial gio OpenOffice.org integration work (i.e. issue 88090 as workspace ucpgio1)

The gio api is a far better fit to the OOo worldview than gnome-vfs is and was pretty straightforward to implement modulo my own bafflement as to why I had an error claiming that mounting share named “X” failed as not existing which incredibly had the root cause that the share named “X” was really called “Y”.

April 30, 2008 06:02 PM

iHealthBeat Discusses Open Health Tools

Last week, iHealthBeat posted an 8 minute audio interview discussing the recently announced Open Health Tools consortium. Open source and healthcare has the potential to change the current healthcare IT landscape, enabling healthcare IT vendors to focus on what truly differentiates systems instead of forcing them to invest in developing basic infrastructure. Collaboratively, OHT will focus on developing free, open source healthcare IT infrastructure for use by anyone.

April 30, 2008 12:09 PM

How much lower can they go?

While I am not surprised that it happened, the fact remains that once a business feels that it's status as a monopoly is threatened, it will do everything it can to ensure that it's status quo is maintained.

From this you can be pretty sure that it was not an isolated instance of corporate bullying. They did the same in Malaysia (the MS MD paid a visit to the IT minister who was in Sabah just after the crucial vote of Disapprove to get it changed to Abstain) as well as similar efforts in Singapore. I am sure similar things were done in South Korea as well.

I was mildly surprised that some local journalists actually bought the MSspeak in that "ooxml is a great thing" and that "the anti-ooxml crowd is really not willing to play fair" was to them accurate. When I explained how MS walked out of the OASIS ODF committees and also that in the ECMA/MS deliberations, no independent third party was allowed to participate does not show that they were or are now any more sincere in working with the open standards people in making a truly and fully specified document standard. It has been a month since the ooxml was bulldozed into the ISO and we have YET to see the final document. How's that for honesty and credibility?

Another low point of that organization is their "new fangled" open source initiative: microsoft. com / opensource. They are championing some individuals as heros for using open source apps and tools on top of a closed source OS supported by closed source development tools. While I am all for a mixed source environment where it makes sense, I note that my challenge to MS to make their tools like their office suite fully supported and available on Fedora is still not taken up. Making it available on Fedora is a guarantee that it will run on other distributions as I personally believe that the freedoms that Fedora adheres to (and possibly exceeds the Debian Social Contract) is a crucial cornerstone in ensure software freedom.

April 30, 2008 08:29 AM

OOXML Questions to Microsoft

Dear Lazyweb,

I have a short list of questions that I would like to see answered by Microsoft. If anyone from Microsoft wants to answer them, please do. If you are also interested in answers, please forward these questions to your Microsoft contact of choice and add the answers you receive as comment.

1. When will which versions of Microsoft Office and Microsoft Works will be updated to deliver a native implementation of the ISO standardized version of the OOXML document format?
2. As Microsoft claims to be a big supporter of Open Standards, when will which versions of Microsoft Office and Microsoft Works be updated to deliver a native implementation of the ISO standardized version of the ODF document format?
3. In the Microsoft Open Specification Promise only the old version of OOXML which is not an ISO standard is mentioned. When will the ISO standardized version of OOXML be added to the list?

I am honestly interested in answers to these questions.

Jan Wildeboer

April 29, 2008 07:40 PM

Provisioning our New Middleware Architecture

Red Hat Information Technology has created a new middleware architecture for our internally-developed applications. Although many of our internal applications take advantage of open source solutions, we have historically used tools such as Tomcat and Perl on a standalone basis and not taken full advantage of our JBoss application stack.

We’ve defined our new middleware architecture based on JBoss technology. This includes an Enterprise Service Bus implementing a Services Oriented Architecture (SOA); Seam, the JBoss integration framework; the JBoss Business Process Modeling suite (JBPM); and Drools, the JBoss rules processing environment.

The entire stack runs on Red Hat Enterprise Linux, taking advantage of the operating system tools for virtualization, provisioning, configuration, and other functionality. The full technical architecture incorporates Cobbler/Koan, JBoss SOA, Xen, LVM, Git, and more.

Architecture Framework

The new middleware architecture is defined as an element within the Zachman framework, in which we consider data, function, network, people, time and motivation mapped against scope, business model, system model, technology model, and detailed presentation. We are excited about the potential of the architecture because it gives us great flexibility while at the same time allowing us to implement projects in alignment with our desired future state. Ultimately, the architecture enables us to reduce complexity while enabling our business users.

Architecture Definition

The internal name for the new architecture is Everest. Among the key underlying elements are the use of the Logical Volume Manager (LVM) to support virtualized guest operating systems and Xen virtualization to assist in the management and deployment of multiple environments.

We use the JBoss Enterprise SOA Platform to provide loose coupling, a messaging infrastructure, process orchestration, protocol translation, adapters, change management, quality of service, quality of protection (encryption and security), and management services. JBoss Seam provides our application framework, unifying and integrating multiple web technologies while reducing complexity for our developers.

Git provide source code management under the new architecture. Git is a flexible, distributed, offline source management tool. It addresses some of the shortcomings of our previous solutions and provides support for parallel development, collaboration and the ability to maintain a consistent deployable branch of the source tree. Git use will result in a smaller, cleaner set of changes to our source code.

Although not strictly part of the run time architecture, Puppet is an important architectural component. Puppet is a configuration management tool that can be used to provide container and system configurations that can be deployed into production without modification. Puppet will also free our Release Engineering team from many manual steps and enable them to move to a review and release mode.

Of course, the entire environment runs on Red Hat Enterprise Linux. We anticipate that some of our developers will use Fedora as an alternative operating system for their development workstations.

Architecture Provisioning

One of the first challenges addressed by the architecture team was the need to provision systems which comply with the new architecture. The deployment implementation supports 20 machine types, ranging from a web-proxy to an ldap service.

The most common types include a host machine, a repo machine, and a JBoss development machine. The host machine type is used to host virtual guests in the Everest environment. It provides a suitable graphical user desktop, automated NFS home directories, a Kerberos login, Xen virtualization, and access to all of the Everest tools.

A repo machine is at the core of Everest. Repo machines are self-contained provisioning, configuration and artifact stores. Repo machines are capable of creating all other machine types, including other repo machines. Features of repo machines include Cobbler for RPM and provisioning, a Puppetmaster for configuration, an Artifactory Maven proxy, Git repos for content under /pub/git, and gitweb running on a known address.

A JBoss development machine is an aggregation of several machine types. (One of the features of the architecture is the ability to consolidate machine types for development and later disaggregate them for the production environment.) Features of the JBoss development machine include the ability to use it on base metal or in a virtual environment; nomachine for remote desktop access; the SUN JDK; an Apache proxy; JBoss Enterprise Application Platform (EAP); JBoss Enterprise Service Bus (ESB); MySQL for JBoss data stores; build tools including Eclipse, Git and Maven; gitweb on a known address; and a Git daemon to serve repos.

Provisioning with Cobbler and Koan

Cobbler is a provisioning and update server that supports deployments via network booting, virtualization, and reinstalls of existing systems. A significant advantage of Cobbler is that it integrates multiple technologies and reduces the need for the user to understand the details of each. The update features of Cobbler integrate yum mirroring and kickstart.

Koan, which is an acronym for “Kickstart over a Network”, is an emerging technology that configures machines from remote Cobbler servers. Koan also enables the use of virtualization and re-installs from a Cobbler server. When launched, it requests install information from the server and creates a virtual guest or replaces the machine it is running on.

We’ve used both these tools as the foundation for deployment of systems complying with the new architecture. In our tests, we can quickly build and deploy a complete development server with no intervention. The provisioning solution significantly reduces both our time to deliver a development solution and to deploy systems in the production environment.

If you’d like to know more about Cobbler and Koan, see the Red Hat Magazine article on Cobbler by Michael DeHaan and the relevant man pages.

Application

We are working with Red Hat engineering, select partners, and the community to share our provisioning vision broadly. Although some elements of the provisioning solution remain as advanced technology and are not guaranteed to become supported elements of Red Hat Enterprise Linux and the JBoss suite, we are confident that the benefits of the solution outweigh that slight risk. We also anticipate that our efforts in IT will facilitate moving the solutions into the mainstream.

More broadly, we are excited regarding the possibilities presented by the new architecture and the capability it gives us to execute our technical strategy. The ability to integrate best of breed applications solutions using a common stack, to project the resulting applications into web environments using common tools, and to take advantage of the scalability and recovery solutions inherent in JBoss will yield substantial benefits.

April 29, 2008 05:29 PM

Fun with an OQO

Just as the novelty of the eeepc was beginning to wear off, today I got a model 2 OQO in the mail. It's a pretty nifty little device. First impression on opening the box and pulling it out was that it was a little 'chunkier' than I was expecting. It's pretty heavy at 3lbs. A whole pound heavier than the Eee. Similar slide-out keyboard to the n810. Much bigger screen.

It comes with Vista pre-installed. I powered it up, just to make sure it's all working. This part took forever. It sat at a "please wait" screen for ages. After 15 minutes, and two reboots, I got to set up my user account. It then spent another five minutes "checking my computers performance" and god alone knows what else. During this time, the device got really hot, and the fans starting running full tilt. For considerable amounts of time whilst I was waiting for it to do something, I was staring at a black screen with a mouse pointer. I had no idea if it had crashed, or was actually doing something. This was my first vista 'experience', but before I'd even gotten to a desktop, I'd decided that modern Linux installations are leaps and bounds ahead in terms of user experience in this regard.

Finally, 25 minutes after I'd hit the on button, I got to a desktop. I moved the mouse pointer, and the screen changed to "shutting down". It rebooted. What was this, the 4th, or 5th time? I'd lost count. After another minute, it had booted up. I fiddled around a little, before quickly becoming bored with it. The fans ran almost constantly. Sitting at an idle desktop, vista pulled around 9 watts, with spikes every few seconds at 13W. Occasionally, it would go as high as 15W. Again, it was completely idle all this time.

After getting bored with trying to beat Vista into using my wireless, I rebooted, and found my way into the bios (Fn-Del). From there, a found numerous things to twiddle (like, enabling PXE as the first boot choice). Surprisingly, there was also a 'Enable ACPI CPU C4" option which was disabled by default. Enabling it didn't cause Vista to use any less power. I guess it's being woken up so frequently that it never gets into those lower states. Given Linux can and will exploit it however, I left it enabled.

Booted up a rawhide install over PXE. Idling, it bounces around 8.4 to 8.9 watts. This isn't an apples to apples comparison though. Sitting in Anaconda is a lot less intensive than an 'idle' desktop.

At this point, I hit my first problem. The ethernet (a Realtek RTL8139, which should be well understood) isn't noticing a link, and hence refuses to get a DHCP release. Some futzing around creating a debug initrd with a shell later, and I discovered that the chip gets discovered just fine. But with one caveat. It detects the MAC address as 00:00:00:00:00:00.

More poking later, but for now, I'm drawing a blank.

April 29, 2008 04:23 PM

Ignoring the MSM (main stream media)

Let me put this out: in the Singapore context, does ignoring the MSM do the country a favour? The MSM, is, by an large, a government spin machine and we all know that. The detainee escape fiasco (confluence of happenstance) and lack of leadership accountability is being pushed aside by calls from a person whom I was proud to acknowledge as my PM, the current Senior Minister Goh Chok Tong. Chok Tong has said that we should move on, with more pressing issues facing the nation. The lack lustre defence of DPM KS Wong by PM HL Lee last week, should not just be swept under the proverbial carpet. Failure of leadership, of accountability, or Doing The Right Thing cannot be forgotten. It might be forgiven, but not forgotten.

I will assume that KS Wong will not stand down and so be it. I cannot but lend my voice and vote to see him out of office. Yes, we have bigger and more challenging things to focus on. I shall continue to ignore MSM - until someone tells me that they have gained credibility.

April 29, 2008 04:01 PM

A wongkanseng moment

I think I just had a wongkanseng moment, you know the ones where you had no idea what was happening, but had oversight of the whole thing and yet, managed to escape any accountability? Boy, I am glad that we have a label for such moments.

In the meantime, a great post by mr wang.

April 29, 2008 03:27 PM

Fiquei rico e não sabia

É. Parece que alguém está enriquecendo com meu nome.

O FolhaInvest, pelo que entendi, é um concurso de investidores. Parece que meu homônimo está indo bem.

Ps.: A dica veio do Edson Tirelli. Valeu!

April 29, 2008 05:09 AM

A whole lot of things to write about…

• Generally I am the last one in the office to move to a new release of Fedora on the main box. I tend to move through stable -> rawhide / development -> newer release on the test box using a variety of combinations of desktop environments and applications. Since the test box is MIA and I was getting tired of the “raah-raah Sulphur” that seems to be echoing from corners of the office, I decided to take the plunge and take the main box to rawhide / development with the intention of moving to Sulphur (stable). Used a LiveUSB to “install-to-hard-disk” activated the livna repo and pulled in a massive upgrade followed by the install of my preferred applications. First impressions - [i] the graphics look way too smooth and beautiful [ii] NetworkManager had some initial hiccups but then began working nicely (UPDATE: Thanks to the poking around the system by mether it turns out that some till_now_undocumented gconf cruft in /home was playing hard to get with NM other than which the disable-enable trick for the connection seems to work nicely) [iii] Pidgin hasn’t crashed yet, so I am off Kopete for a while [iv] took some time getting used to the packagekitd + yum thing with yum getting locked now and then. I generally turn of yumupdatesd and then do yum install / remove etc and guess getting used to PackageKit will take time [v] scim and Bengali (India) support work nicely [vi] fonts on the browser render well, I could do with some of my browser add-ons though, but I guess they will come in due course [vii] the firstboot doesn’t include the SELinux bits so you can trip up getting the system installed and booted into SELinux “Enforcing” mode. Fun, fun, fun
• The experience with the local mirrors viz the GLUG-NITH and the WBUT ones hasn’t been a happy one. The former times out rejects the connection (as per conversation with Debarshi) with an unerring frequency while the latter doesn’t mirror development yet. The usage statistics for the GLUG-NITH mirror are here, and it does prove that more mirrors and torrent seeds are required. If you know someone who can help in getting this done, do get in touch with the Fedora folks in India via the fedora-india mailing list or the #fedora-india channel on FreeNode. In fact, if you want to help out with the FreeMedia project in India, these would be the means to get in touch. At the last count there are more OPEN media requests than we can handle and we would really love to see more contributions for this effort. Getting the OS bits to consumers and contributors go a long way in ensuring that the latest and the shiny greatest release is available for testing and contribution.
• The list of projects this time looks amazing. Here’s to a wonderful time hacking and building communities. A slight note on the insanity that prevailed for a short while with various LUGs pushing out lists like “here are the number of folks from ‘our’ LUG who are participating as candidates or mentors”. Without being too sarcastic, that reminds me of the time a number of coaching centers take out advertisements which proclaim to the world and its uncle that “xy number of students appeared in the top ranks for such and such common entrance examinations”. It isn’t really about the breast beating - it is about ensuring that the mentors and the candidates work in harmony and help ensure that a community of usage is built around the project. If they have a collective of people who think, work and learn together it is then that the massive push like the GSoC would be really effective. Till then it would be like the annual festival which is slowly losing some of the sheen.
• Thanks to all the fine folks at Nokia I finally have my n810 at hand (thanks to ramkrsna from getting it down during his 17 flight backpack across the US trip). First impressions - [i] xrandr breaks a whole lot of things (should not have tried that really) [ii] scim stuff is not suited to a tap style of input [iii] I miss the fact that the button is not outside and I have to expose the keyboard to use it [iv] the keyboard seems a bit flaky [v] a lot of my prejudices around the on-screen tap based usage would require a re-visit [vi] the application wish list is an interesting document.
• Clare has a lovely welcome note for GSoC candidates that can apply to all the other projects too.

And in some other news

• Yesterday evening we had a lovely meal at Barbeque Nation which is doing a soft launch at Pune. The concept of a “live barbeque” at the table is fairly unique for me at least and the food was lovely. For a change the music wasn’t really too loud, the seating comfortable and we had a whole lot of stories to share over the fixed price 5 vegetarian, 5 non-vegetarian “eat all you can” barbeque feast. The food could have done with a touch of less salt and I really wished I had dragged my camera along.
• Runa is back from her vacation and she re-arranged the bookshelf and I have space for more books. She also brought some goodies (read palm-candy) for me. Yay !!
• During an unguarded moment of enthusiasm I had opined to Lucas that the RoadMap needs to be “un-geeked”. So, I plan to sit down during the coming weeks and ask the lamest of questions on the now-released-roadmap and be a general pain in the posterior for all the fine folks who responded to the roadmap call.
• ps. we are looking out for a software engineer.