Getting ready for IT360 / Linuxworld

When I have been asked to present, or had something I submitted accepted there is jubilation. I am getting ready to embark on a new experience. Every time I give this lab it is a new experience. New people attend, bringing their own unique situations and backgrounds. It is not just the people that are different each time. The conferences each have their own flavors. The current hot button issues and general industry excitement changes. Linux on the desktop here, Virtualization there. Dual core and performance per watt here, which RDB runs fastest there.

Right now there is a lot of buzz around all the meetings MS is having with Linux vendors. The SUSE / MS deal. The way SCO's lawsuit is going. Redhat showing up and joining an MS interoperability standards group. A founder of Debian power lunching at Redmond. It would seem the subject of Linux and MS Windows interoperability is not only germane right now, but may be in flux. Then there is the thing about Dell customers wanting Linux desktop options.

That is the challenge of the lab. Keeping it fresh, new, and up-to-date.

The technology of the lab evolves: I started doing this by literally carrying around a full bore 1U server with me! It was backed up by a regular size desktop. And I had to have a monitor, and there were no flat panels in stock at the time. The shipping crate was huge.

VMware and laptops catching up to some degree to desktops as far as memory and CPU speeds changed that math. Now instead of shipping a server, I stand in line at the security checkpoint pulling laptop after laptop out of my carry on bags, so they can be X-rayed. The people behind me really hate that part. Like a geek clown car or something is going on ahead of them, and they are of course running late for something.

And of course, Linux isn't standing still either. More on that in a sec.

The main point here is that I'm jazzed. Ready to go. Ready to talk about the new things on everyones mind, and to see what people are up to with the technology.

Part Two

I realize what I have done. What was I thinking?

If the last time I gave the lab was recent, it is not so bad, but when it has been a while, then technology has had a long time to drift upwards. Ok. Rocket upwards. No one wants to come to a lab using yesterdays Linux. They want to know how this will all work *today*. Or tomorrow, when they go home and try it in their place. They will have the LiveCD, and the lab book. They should be able to check it out at their place, as long as it is not the NSA or someplace like that where if you brought a bootable OS disk in they would probably have to have a long talk with you.

In the first version of the lab, I chose Knoppix 3.2 as my student distro. It had all the packages I needed, and I ran through all the computers at the office I could booting it to be sure it worked with most everything. I had no idea what the attendees would bring to the lab. One time a person brought a computer they pulled out of the IT junk bin, and did not see if it would boot. It turned out it would boot the internal hard drive, and run Windows 95 or 98 or some such, but the CD drive was trashed, and there was nothing I could do but put them with someone else who had a booting computer. That appeared to actually work out pretty well though, as they helped each other through the lab.

Just to be safe, I borrowed a few spare laptops from our desktop support manager and brought them the next time I gave the lab. The idea was obvious: I could pre-certify that the laptops were Linux compatible. the hard drives were blanked. All I cared about were the CD's being bootable. They were stolen out of the storage area I had them in after the lab. The manager of desktop support at BMC was not happy with me. One of them was his *favorite* laptop, and IBM Thinkpad! Ouch.

Updating the boot distro means re-certifying everything in the lab still works. Re-screen capturing everything. Changing the verbiage of the various sections to match new behaviors of various packages. A 122 page document basically has to be re-edited every time to match the current reality of the lab, and the state of the art 'out there'. You can also start to develop some pretty firm opinions about good and bad look-n-feel changes at times like this. If I have to re-capture and re-document because someone decided they wanted to move something from one place to another and I can not see any reason for the change, it makes me grumpy. I am not saying all change makes me grumpy. If the menus were poorly organized before, and now they are better, that makes me happy. It will make the lab much easier to give. It is nice to not have to answer questions like "Why in the world did they do it *that* way?". Think of all the scorn a 13 year old can put into the word “that” and you get the idea.

I usually have to answer questions like that by relating them back to sunspot activity.

“Why in the world did they put this menu option here, and that menu option there when the two things are related and should be together?”

“Err.. well. You'll recall, this was developed last year, when there was all radiation coming from the Sun during Solar Max... yeah. OK. I'm not buying it either. Sorry. I got nothin'”

Then there is the printing and the CD manufacture, and the downloading of OpenCD (4.0 now out) and burning those (I always give away a copy of whatever the current OpenCD is at the lab), and figuring out how to get all that to the show ahead of me.

There is also some basic BSM logic at work here:

Any change increases risk. That is why you have a Change Manager. I passed my ITIL test. I'm Foundation certified.

At the lab I'll have about one shot to get it all there and working. An entire mini-company I carry around, plus all the 'users'. This fear of a single point of failure is why I just bought the new Acer to take to the show. The IBM T41 has been a total trooper, but it is now over three years old and has been through the mill. I always had it backed up with my eMachines 5312, but it is deceased. As they say:

"Failures happen, and absolute failures fail absolutely."

Or something like that. In any case the death of the 5312, which was younger than the T41 has made me realize that the T41 will not be with me forever. The chances it would go to meet its maker increases by the number of both airline miles traveled and number of people in the session.

Part Three: The Denouement

Part three starts when I get to the show, and get to the room, and see the space. I don't know why, but wondering around and getting a feel for the room before anyone is in it helps reduce stress. Everything I need for the lab is in my hotel room, I have assembled it to be sure it is all still working, and I have run through the lab. Last time I gave the lab in Baltimore, I realized I was missing a power cord to the DHCP server, and had to make a run to Best Buy to fix that. But I did, it worked, and I was starting to have hope.

Once I see it all got there: all the bits that make up the lab, and it working, I start to relax. The folks at the show let you know who to call for problems, who to deal with on power and the network and whatnot. I have hooked up a computer to the projector to see which one will work the best for that (that is always odd to test: clearly there is no standard for digital projectors yet. Sometime Linux works best, and most of the time the Mac works. Last time it was Linux: the Mac just could not deal with something it was getting signal-wide from the projector. I thave talked to the show network folks to make sure they know that if they hook the lab to any other network, they will crash everything. It happened the first time I gave the lab. I know it will.

Folks arrive, we talk, we get started, and the next three hours fly by. At the end I realize why I went through it all: I like to do this. Call me crazy, but this is actually fun and interesting.

Back to Part Two: Changes

From the arc described above, I am currently at the beginning of part two for this go-round. I am trying to figure out what to change. What to update. What to leave alone. Who is my likely audience, and what will they want to walk away with for their three hours of time invested?

After some pacing about, I decided to start with looking at what the state of the art of a couple LiveCD's might be. In past I have used Knoppix or PCLinux for the lab. The last time I used Knoppix 3.6, and went through a process with my often collaborator Sam Stengler to build a custom, CD version of it. K36 had dropped Evolution from the LiveCD and I needed it back. That meant we had to take a bunch of stuff off the default Knoppix LiveCD disks to make room for Evolution.

You don't just do that to the LiveCD. You have to roll it out to disk, un-compress it. chroot to it, make changes, add and delete, guess when you are close to having it back to the right uncompressed size such that it will roll back up, re-compress and assemble the LiveCD bits, and then test it. It was not the work of one evening.

I had new new respect for those that build these things in the first place. I do not want to do that again. On the other hand I do want to update to newer apps and kernels than what is on Knoppix 3.6. My first decision for a change to the lab is to move to a BootDVD. Most every PC I have seen recently, especially a laptop, has a DVD drive on it. So I decided to make that a requirement of the lab. That is a big change: we'll see how that works out!

Recent work with Ubuntu has me thinking that maybe Kubuntu would be a good option to test, so I grabbed the most recent boot DVD of that.

Why Kubuntu rather than Ubuntu one might ask. In my opinion KDE is a better desktop for people who are coming from MS Windows desktops. That is core to some basic assumptions of the lab design: that the people in it:

Nothing against Gnome: I use it all the time these days. In the lab I do not want to deal with more learning curve than I have to, and I would rather focus the learning on the servers and the apps and the protocols, not desktop navigation.

And KDE has lovely plumage (Sorry. Been watching Monty Python DVD's with my daugher). That is to say, it is very attractive to look at, and has some nice effects, all for a relatively small hardware cost (not even counting Vista ....)

Just out to also consider is Knoppix 5.1.1. Also KDE based. Knoppix has served very well in the past, even if I had to modify it. I'll get a BitTorrent going on that too.

Neither OpenSUSE nor Fedora Core have LiveDVD versions of which I am aware, so I am going to start with these two distros. As soon as they come down, I'll sanity check this by booting them on every machine I can lay my hands on. They do me no good if they do not have great hardware recognition. Ubuntu 6.10 had no problem with the Acer 5610, so I am hopeful here.

The other big change to the lab so far is the Acer laptop I mentioned in my last post. Currently it is running Ubuntu 6.10. VMware is up and running already, and works great. The laptop actually suspends and resumes like it is a Mac! It is working like a champ so far, so I am hopeful. The T41 still has OpenSUSE on it, so I'll be able to show both an RPM and a DPKG based distro to anyone that wants to compare them.

(Update: Hey.. look there. A new version of Mint, 2.2, just came out. )

Utterly new for this version of the lab will (hopefully) be the Mac as a server. In past I had the iBook, with a copy of the OpenOffice Impress slides on it, and that was what I usually drove the introduction bit from. But Parallels and the power of the MacBook changes the story of OS.X for me, as it related to the topics at hand: OS.X is now a viable enterprise desktop too, even if the exact applications are different, or ports from Linux. The Macbook also adds a new capability I have never had before: three copies of the lab servers! Call me crazy, but I like the idea of that new redundancy. Parallels is now supposed to be able to import the VMware disk images, so I'll be checking that out soon.

My main worry right now: the show projector. Not a single computer I am bringing runs 1024x768 as a default resolution. I could bring the IBM X30, but that would really torque off the people behind me at the airport. Four computers at the security checkpoint!! 1024x768 is normally the maximum resolution most of the projectors I have seen at shows. The Apple usually deals with such things really well, but last time it didn't.

Time to pace and think....