Some of the sessions to highlight:

• “A day in the life of Facebook operations” – Gathered some info on what happens when you scale bigger than pretty much everyone else in the world.
• “Datacenter Infrastructure Innovation” by James Hamilton – I’ve seen a video of an older version of his presentation, but still some new info coming in about stupid things we’ve been doing for years in our datacenters.
• “In the belly of the whale: Operations at Twitter” – similar to the Facebook presentation, but the name of one of their monitoring tools was worth the mention: “Whale Watcher”

There were also a number of sessions on different performance tools. Hopefully posts on some tested out for real are soon to come.

On the slightly less technical side, there were some great talks about culture, including quite a few talks on DevOps (more on that in a later post):

• “Creating Cultural Change” by John Rauser – some great examples on how to create change of culture within your organization by trickery, humor, and other methods
• “Ops Meta-Metrics: The currency you use to pay for change” by John Allspaw – was a great talk that could also be a blog post of its own. Basically, as operations people, we tend to be afraid of change because we associate it with downtime. To get around that, we change more often and track just how often things really do go wrong.
• “The Upside of Downtime: How to Turn a Disaster Into an Opportunity” – Talk on how to use communication and transparency from downtime to create trust from your customers.
• “The Mobile Web: The next frontier” – presentation by Strangeloop on the monetary value of performance

Another great set of videos from the conference is the “Choose Your Own Adventure” talks with Adam Jacob from Opscode. You can head over to one of the videos and see them all listed in the related videos. I didn’t get to that session, but watched all the videos and wish I had now .

It was a great few days and was followed by DevOpsDay USA over at the LinkedIn campus, more on that to come!

Anyway, so I started trying to find a nice piece of software to publish my pictures fairly easily from iPhoto. Since I would rather have a CMS (but was willing to put up with the likes of iWeb for organization integration), I decided to take a look at a few open source products to see if one of them would fit the bill. I found an awesome combination that I thought was worth mentioning.

I found the iPhoto2Gallery plugin for iPhoto. The website says it works with iPhoto 08 but it seemed to work just fine for me on iPhoto 09. After I setup a Gallery2 site quick and the plugin didn’t seem to login. Then, when searching for a solution for that problem I hit another lucky break, I ran across this module for Drupal to emulate the Gallery API for publishing! Once the module is enabled, I tried it out and it seems to work flawlessly!

If you’re looking for a slick way to publish your photos from iPhoto to something that’s not Flickr or Facebook and are not wanting to pay for a .Mac account, check out this combination of software. It’s pretty slick!

Just as a note, the default Drupal gallery module is ‘ok’. It’s much better if you add the previous/next buttons

For myself, I wanted to know if WordPress would run faster on an nginx or Apache server. There are plenty of how-tos out there on setting up nginx to use fastcgi for PHP applications, so I won’t go into that, but I happened to use this one.

For my little test, I used Apache Bench (ab) on a separate machine attached to the same switch. I took four tests and averaged the total time to complete the requests given by the output of Apache Bench. Below is a test of 100 requests one at a time (total time in seconds, lower is better):

That wasn’t at all what I was expecting. It wasn’t any different at all, really. The numbers were: 40.00 seconds for nginx and 40.04 for Apache. Add a little roundoff error in there and we really can’t say much about the results.

The numbers get a little more interesting when I start adding a little concurrency:

Definitely a trend, but even at 40 concurrent connections it’s not really anything worth writing home about. With a little tweaking I’m sure the concurrency issue can be throw into a whole ‘nother direction, but I just took what came out of the “box”.

Another interesting thing that I noticed was the memory usage between the two. With Apache, the web server used 23400K of memory. nginx used significantly less memory than that weighing in at a measly 4356K. However, since it can’t interpret PHP on its own and uses fastcgi, we have to add that in as well. That adds 19228K of memory, totaling 23584K, slightly more than Apache!

Apache and nginx seem to be almost the same when being used to run straight PHP applications. If you’re looking for a lighter-weight straight application server for PHP, I probably wouldn’t look any further than the LAMP stack since it has been made extremely easy to install and configure on popular Linux distros. Start adding load balancing, web proxies, mail proxies, and fault tolerance and then I’d start looking at nginx. Otherwise, I’ll be sticking with Apache for my PHP apps for now.

When I got up and running, I went to install the 2X client again:
wyatt@host:~$ sudo dpkg --install 2XClient.deb
[sudo] password for walterw:
dpkg: error processing 2XClient.deb (--install):
package architecture (i386) does not match system (amd64)
Errors were encountered while processing:
2XClient.deb


Wrong architecture. Dang.

Not to worry! 2X also distributes the binaries in a tarball so, using the 32-bit libraries, one can just run the binaries.

First, download the tarball. You can download it using your browser from their downloads page, or from the terminal via:
wget http://www.2x.com/downloads/AppServer-LoadBalancer/2XClient.tar.bz2

Untar it:
tar jxf 2XClient.tar.bz2

Copy the contents to /opt:
sudo cp opt/2X /opt/

Install the ia32libs:
sudo apt-get install ia32-libs

You can then create a launcher within the Gnome menu or whatever desktop manager you want.

To create the launcher, the command to start the client is:
/opt/2X/Client/bin/2XClient

Create 2X Client Launcher

That’s it! The 2X client should launch and run beautifully.

As Matt Asay points out in Microsoft’s strategy in the virtualization market, ‘good enough’ and convenience sometimes are plenty to take over a market. Microsoft has something that Google and Zoho do not: desktop operating system market share. Just like Microsoft has done in the browser market to take over most of the market until Firefox came around (they still hold the majority), they essentially have a ‘botnet’ of users in Windows users. If the integration with Windows 7 and Office online is ‘good enough’ and integrated in the OS already, users may be drawn to it out of simplicity. Often consumers don’t care about what technology is better. In fact, I would argue that most of the time they don’t care about what’s better. They just care about what is easily available and the quickest. If Office online can meet their needs in a pinch, they probably won’t explore other options.

On the flip side of that, Google and Zoho do have a pretty good foothold in the web-based docs market. Google Docs or Zoho just might be ‘good enough’ to keep users already using their services. Plus, services like these often don’t integrate with others, so they tend to be somewhat viral in nature since collaboration with others requires those users to sign up for those services. As OS X and Linux start to become more popular alternatives to Windows, Microsoft’s advantages in integrating its OS may start to diminish, but their marketshare is still pretty high.

We still have to see exactly what Microsoft has up its sleeves for features in Office Web. Microsoft is calling their Office Web apps an “online companion” to its desktop applications, so we will see what they offer in features. Promised features so far are the ability to create documents and do basic editing, something that both Google and Zoho can blow out of the water. Only time will tell what will happen, but Google and Zoho seem to have finally brought Microsoft out of its fortified position on the ground with Office to do battle in the clouds.

If you take a look at the minimum system requirements, Windows Vista Premium requires 1GB of RAM. Windows Vista Basic requires 512MB of RAM, but that doesn’t suit most users’ needs. OS X Leopard, on the other hand, only requires 512MB. (Can we call it OS X Leopard Premium?) Now, I’ve run Leopard on 512MB of RAM. The system is usable. Not incredible, but not too bad. Ever try running Windows Vista on 1GB of RAM? Pretty much the same. It works. It’s just not overly usable. I guess I am not quite sure how the PC that comes with 3 times the required memory is a better value than the Mac with 4 times its required memory. The fact that to get a well-performing Vista laptop, you need more system resources is a selling point?? Since when? What are your thoughts? Out of control ad campaign or legit?

xVM is comes out of of the Xen project which was acquired in 2007 by Citrix.  Although still available as an open source project, XenSource wraps a nice set of Windows based management tools around Xen improving management, and increasing the cost ofcourse.

It will be interesting to see what the management tools will look like when xVM is released, as it’s $500 per server price point (for support) is intriguing as it would be an efficiently priced, commercially supported option for server visualization.  Also, with Sun’s support of multiple platforms, it will be interesting to see how cross platform the management tools will be as many tools are Windows only or only partially function on other operating systems.

Additional features of xVM include support for SPARC and x86 processors, multiple guest OS’s, live migrations, and a Java / browser based KVM. Currently, xVM is available via source code.  The rest of us mere mortals will have to wait for the binary release.

Sun has also received good reviews of its desktop virtualization product, VirtualBox.  MacWorld recently posted a video reviewing the product.

I was really frustrated when the software didn’t react as it used to. For example, I use Webmin to a great extent both personally and at work. Whenever I wanted to webmin.com, I’d start typing ‘webmin’ in the address bar and once my history started popping up, I’d hit the down arrow, hit enter and I was there. Now, when I do the same thing I get taken to my Webmin login pages since “Webmin” is in the title of the page and I tend to go there first. That change frustrated me to no end until one day I wanted to get logged into one of my systems. Absent-mindedly, I started typing in “webmin” into the address bar, hit the down arrow and hit enter. Without even realizing it, I was brought to the place I wanted to go without typing in the address. Now, it comes second nature. In my work I use the Zimbra wiki a lot and tend to go to a lot of the same pages for reference, and I can just type in “wiki Zimbra” and then the topic I want to find the pages very easily. Even though Firefox 3 is still fairly new, I don’t know what I’d do without that feature anymore. Mozilla did a great job of giving the me what I really needed, not what I thought I needed.

The supported “Network Edition” is still in beta with limited support, but the community packages can be downloaded here.

According to the website, you can currently purchase computers with Linspire pre-installed from a large amount of major retailers including Wal-Mart, Buy.com, and Amazon. Although the timing is interesting since the acquisition happened such a short time ago, the changes should allow Xandros to focus more on doing one (or two) thing[s] well rather than having to maintain three different distributions. Xandros seems to be learning a lesson from Novell and Red Hat, using Freespire as its community-based testing and proving grounds for its pay-for, supported Linux distribution.