Next up in the series of LISA ‘08 guest bloggers: Kyrre Begnum and Æleen Frisch with Virtualization and Virtual Worlds: A Conversation

Æleen:
One of the most fascinating features of virtual machines, we find, is live migration, allowing a running virtual machine to travel from one host to another without the user even needing to know about it. Normally, the new server would be right next to the old ones, so in terms of physical distance, the VM would not go far. But it could if it wanted to.

Kyrre:
Last year, a collaboration between Exponential Consulting and Oslo University College built a framework which enabled virtual machines to travel over the Atlantic ocean. We also included the college of Gjøvik,located about 200 km north of Oslo. Much closer and only forests separate us, but the virtual machine could live migrate in very few seconds. Why did we do it? Because we could and because we did not know what awaited us when we were done.

These crazy projects make me daydream. At the college I teach, do research, and manage over 200 systems. Only 11 of those are physical; the rest are virtual machines. The users of my virtual machines have the typical user problems: they set wrong permissions somewhere and cannot log in anymore or forget their passwords. That’s ok. But while I am watching a virtual machine jump across the Atlantic, I fantasize how users knock on my door and tell me about their virtual machines now being somewere in Hawaii, because that’s where CPU time was cheap right now. They talk about how they programmed the virtual machines to flock like birds to the country where there are the most current users to their service. Stories circulate how one colleague made his virtual machine over-anxious so that when it was the target of a port scan, it got scared and ran away. It has been lost ever since.

I know, this is far away, but the world will get there. The first step is to master the technology and learn how the flexibility of virtual machines can be utilized by us so that we can deploy our systems effectively. Next, management of virtual machines will become transparent and adaptable to everyone, even the users of them. Then, organizations will cooperate, and markets will be built enabling virtual machines to travel the world. When that happens, my tasks at the college will include one more item: sending out emails entitled “Have you seen this VM?”

Æleen:
I’ll hate it when my VMs get a better vacation than I do!

We have written this tutorial because we want inspire others and given them information about how they can use virtual machines in their organizations. Virtual machines can be used in many scenarios. In our course, we will show solutions to interesting problems. If you attend, you will learn about what virtualization can–and cannot–do and also get plenty of information to more than get started using Xen or VMware. We hope to see you there.

S1: Virtualization: What is it Good For?

Kyrre Begnum is currently completing his Ph.D. in Network and System Administration at Oslo University College in Norway. Part of his research focuses on managing virtual infrastructures, and he is the author of the Manage Large Networks (MLN) VM administrative tool

Æleen Frisch has been working as a system administrator for over 20 years. She currently looks after a pathologically heterogeneous network of UNIX and Windows systems. She is the author of several books, including Essential System Administration (now in its 3rd edition) and the SAGE Short Topics booklet A System Engineer’s Guide to Host Configuration and Maintenance Using Cfengine, coauthored with Mark Burgess. Æleen was the program committee chair for LISA ‘03 and is a frequent presenter at USENIX and SAGE events, as well as presenting classes for universities and corporations worldwide.

Next up in the series of LISA ‘08 guest blogs: Rik Farrow.

Here’s what Rik has to say:

I will be wearing two hats at LISA this year: one as a tutorial instructor, and the other as ;login: editor. My editor’s hat, consisting of an unruly fringe of hair, will make it plain that I am on the prowl, looking for articles for 2009. I already have a good lineup of articles for the February 2009 issue of ;login:, but want to find authors who can cover the big ideas in sysadmin for the rest of the year.

Two things that come to mind immediately are VM and configuration management. The IT world has gone crazy for VMs, and that changes the nature of sysadmin. Rebooting one machine can mean the halting and restarting of a dozen VMs, for example. In a sense, the configuration of VMs is not that dissimilar to the configuration management of separate systems, rebooting aside. And VMs themselves are still a work in progress. Exactly how much swap do you devote to a VM, and should this swap reside in a virtual disk or a swap partition? I have yet to hear a good answer to this question.

And not that the configuration ‘wars’ are over. There are still at least five contenders in the ring for dominance in the configuration management arena. I imagine that there will be BOFS as well as fierce
hallway (and poolside) discussions about whose software works the best. As often is the case, there will not be a definitive winner here, not just on merit, but because people have a lot of time and brainpower tied up in a particular solution.

I’ll also be watching the various pre-LISA workshops. I’ve found that sitting in on workshops (and reading the summaries) provide fertile ground for future articles. While most workshops focus on current issues, I always learn something new that I should be paying attention to. I notice that ZFS gets its own workshop this year, a hint that either ZFS administration is still new and complex, or that this area is just too profitable to ignore. I wonder why this is a workshop and not a tutorial?

During Tech sessions, I see lots of old friends as well as many new faces in both the papers track and the two tracks of IT sessions. I found myself thinking “oh boy” several times as I read through the lineup. This week will be fun, including the hallway track, where I can get some face time with people.

I will also be wearing my NSA hat. Of course, you can’t see my NSA hat, because it is classified (and invisible as well), but I am determined that I will help people who attend my tutorial re-enable enforcing mode in SELinux. I’ve worked with SELinux over the years as a user, and often as not been frustrated by it. If an application didn’t work, I just put SELinux into permissive mode, and then, likely as not, forgot about it after I solved whatever problem I was having. I decided it was high time I learned more about SELinux.

Red Hat, and other distros, continuing insistence on including SELinux features also piqued my interest. SELinux must be worthwhile or it would have been dropped a long time ago. And second, usability features
have appeared, such as setroubleshootd, sealert, modules, and booleans, that have made managing SELinux much simpler.

Honestly, I never got an NSA hat, even after over a dozen visits. The RCMP gave me a nifty hat, though, so perhaps I will wear that one.

As always, I look forward to going to San Diego, and hope to see you there.

Rik Farrow has been teaching UNIX security classes since 1987. He wrote the second book on UNIX security, as well as hundreds of security-related articles. His experience with Linux security goes back over ten years and has led him to believe that sandboxing applications with SELinux is not just a good idea, but necessary. Rik Farrow is also editor of ;login:.

Find out more about Rik’s class Working with SELinux (New)