Category Archives: TOS

Success at running images!

Towards the final weeks of the semester we actually managed to put together a cloud that was basically running. I say basically running because there still seems to be a few bugs here and there that causes some issues every once in a while.

In the week that we managed to finally get an image up and running Dave actually had decided to reinstall eucalyptus on our cloud with the packages that Eucalyptus team had end up putting together halfway through the semester. During class he actually managed to quickly install all of the packages and set everything up, while long and I managed to package and image with a kernel that was working on the ECC cloud, and we actually managed to get it running.

This was a success at running an image and all seemed well but there was still some work to be done, we could run one image but when we tried to run more it was not allowing us to due to not enough resources and seemed not to be assigning IP’s correctly. So there must have been a problem with the network configuration, Mike was tasked with flattening the network and after some tweeks that Dave did to the networking configuration on the cloud all seemed much better.

We were able to get multiple instances up in the long run and the troubleshooting that Dhimitri and I have been doing seemed to help locating where some potential problems were by going through the nclog file.

From the blog armindoa » WSU CS by mindoftw and used with permission of the author. All other rights reserved by the author.

Creating a bootable disk to flash BIOS

One of the issues that we had while setting up our cloud was incompatible BIOS revision on the machines we were using. They were out of date and did not allow the use of virtualization, so we had to update the BIOS a certain way with the DELL Executable that was the only way to update the BIOS was using an EXE.

 

The first step to creating the bootable disk was downloading FreeDOS so that we can boot up into FreeDOS and run the executable.

In order to do this you do: wget http://www.fdos.org/bootdisks/autogen/FDOEM.144.gz

Then, gunzip FDOEM.144.gz to unzip the file.

 

The second step is to copy the BIOS flash utility and the BIOS image that needs to be used to upgrade, and mount it to a floppy disk image. Here is what you do in order to complete this step:

modprobe vfat

modprobe loop

mkdir /tmp/floppy

mount -t vfat -o loop FDOEM.144 /tmp/floppy

After mounting the Floppy you want to copy the EXE. that you downloaded for the BIOS:

cp DELLBIOSVERSION.exe /tmp/floppy (not actual name of the bios)

then unmount the floppy: umount /tmp/floppy

 

Step 3 is to burn the bootable CD which emulates a floppy device.

mkisofs -o bootcd.iso -b FDOEM.144 FDOEM.144
cdrecord -v bootcd.iso

After you have created this bootable disk, all you have to do is boot the CD in the machine and basically type in the EXE. and it will autoflash to the new bios.

From the blog armindoa » WSU CS by mindoftw and used with permission of the author. All other rights reserved by the author.

Final Thoughts about CS401/Eucalyptus

I had a lot of fun working in this class it was definitely a difference experience than almost all my CS courses before hand.   I liked working on a big project my only request for the future is to see more code but, something that was really unavoidable for a beta-test style class that we had.  I know it wasn’t anyone’s fault just too many problems with the cluster to really get into it.  Just because we didn’t get coding done doesn’t mean I didn’t learn quite a bit about how exactly working on a big project was like.  Learning to use git, irc, and being on a list server were pretty new experiences for me, aside from gaming I had never used irc, git seems to be invaluable to know since every employer I’ve talked to seemed excited that I already know it.

 

Main thing I’ll take away from the course is exactly what cloud computing is, all the nomenclature we’ve learned, how to work on large scale projects, how to communicate with a large group of people from different time zones and different backgrounds in programming/computer science.

 

Final presentations today were great, seemed like every group found a lot of different topics to talk about and cover and they all seemed to master their specific topic.

From the blog jamescelona » WSU CS by jamescelona and used with permission of the author. All other rights reserved by the author.

Final Wiki Adjustments

I never got around to putting up this blog post last week but I figure it’s worth posting.

 

After siting down with Chris while he worked his way through our wiki was great, really gave me a chance to see how a wiki flows through another set of eyes.  Luckily we’ve all been doing a great job of working with the wiki adding the needed information and adding it coherently.  Only issues really were fixing formatting adding things like, “don’t forget to use sudo” to the top of commands rather than after.  Only “big” changes I made to the wiki was adding a bit on git and how to install python user a package manager.

 

I enjoyed working with Chris getting his input, I think it was smart to do it on my own first, since I knew a bit on how to get everything running from doing it a few times, then having him go through.  Since I never posted this Karl mentioned that there is a whole field of engineering in something like that, actually testing your code/wiki/idea on an audience and jotting down notes.

From the blog jamescelona » WSU CS by jamescelona and used with permission of the author. All other rights reserved by the author.

Week of April 30th 2012

This week was the last in-class meeting of the semester.

We discussed various ideas about how we should use our Eucalyptus cloud in the future. One of the ideas was to provide students in lower level courses with images of the cloud that supplied them with all the tools they need to do their work.

We then were able to meet up with our groups and further discuss completing our project.

I worked on creating a Power Point presentation that represented a visual summary of our research on Eucalyptus’ Architectural Overview as well as general Cloud Computing information.

Also, on Saturday we met up on campus as a group to do a practice run of our presentation. After a few run-throughs we felt pretty good about what we had prepared.

I decided on a topic and wrote my Ethical Analysis regarding Eucalyptus. This was a learning experience as I began to do my research on security issues with Cloud Computing/Eucalyptus. I had previously heard about the issues, but upon doing some research on the topic it furthered my knowledge of things I wasn’t aware of.

From the blog nzahid » WSU CS by nzahid and used with permission of the author. All other rights reserved by the author.

Final Presentation

This week I spent most of my time figuring out what was left to finish in our eutester repository. I also used a lot of screenshots from github for my powerpoint presentation. I worked with Nate, Coady, and Marcos to put together our final project. We all put together screenshots and code samples from the changes we added to the repo to show everything we accomplished this semester.

Also, I wrote my paper on the Eucalyptus Ethical Analysis this week. It was interesting to look at the software and find out where it was good and where it was not so good. As it turns out, Eucalyptus is not 100% free and open source. It is better described as “open-core” because it only allows certain aspects of the product to be used for free. To add on more powerful features, users must pay a fee, which goes against the entire open source mindset. Users are only given a free trial of the software and all of the products that can be used with Eucalyptus are proprietary. There are other options that are similar to Eucalyptus, that are actually free and open source, but it appears that Eucalyptus is more mainstream and user friendly. 

From the blog trevorhodde by Trevor and used with permission of the author. All other rights reserved by the author.

Test Run of the Wiki

This week I am going to start from scratch and try to install Eutester+euca2ools from our wiki only, keeping notes along the way to see if we need to update anything to make it more clear.  I imagine from what I’ve been reading that we’ve made quite a large amount of progress over this course and when I had to restart last month it took me maybe an hour? (would be been less if I realized >>> meant that you type python ) but that’s besides the point.  This blog post will serve as a note pad for anything our wiki should make clearer.

 

 

Updated this:

I decided that I would find someone I could use that doesn’t really know much about Eucalyptus and at least knows how to work Linux.  It turned out pretty well, I have some notes jotted down that I can finalize the wiki with to make it as fast as possible.  Everything was pretty well laid out, aside form small things like the note “don’t forget to use sudo” after they entered all the commands.  One thing I need to add is that we never directly say to install boto.

 

Overall this test went pretty well.

From the blog jamescelona » WSU CS by jamescelona and used with permission of the author. All other rights reserved by the author.

Week of April 23 2012

This week we got together in class and decided on how we would go about presenting our portion of the Eucalyptus project. Our group, because it discusses the Architecture Overview, will be the first group to present on May 9th, 2012. We decided it is probably a good idea to do a couple of run-throughs of our oral presentation so that we are prepared on the day of so we met up on campus on Saturday to get started.

After meeting on Saturday, we came up with the details regarding what our presentation will consist of.

I added some finishing touches to what I already have written on the storage controller and machine images.

We also received the requirements/credentials for the Ethical Analysis of Eucalyptus so I began working on that as well. The topic I chose to discuss was security and cloud computing/Eucalyptus. There are security issues with cloud computing in general so I thought that would be an interesting topic to tackle.

From the blog nzahid » WSU CS by nzahid and used with permission of the author. All other rights reserved by the author.

Ethical Components Regarding Eucalyptus

Eucalyptus has a lot of vulnerability when it comes to ethical assessment.  Eucalyptus puts the power of a cloud at the keyboard of its users.  It is open-source and the community could potentially use it with malicious intent.  Who can be held accountable if the software is used for destructive purposes?

Eucalyptus has a very broad terms of use policy.  It simply states that contributors and users are responsible for the contributions they make toward the project and the implementation of the software. Users agree that it will only be used for lawful purposes.  Though, in the technology world, there is a lot of grey area. Laws often vary from state-to-state and even more so on the international level.  What if a cloud is hosted in one state, where the laws are loose, and is utilized in another state where they are very tight?  What legislature applies?  This area is still very undefined, and proposes great risk to contributors, users, and Eucalyptus’ responsibility.

On the other hand, the more detailed the terms of use becomes, the less versatility and flexibility you have with the software.  You begin to remove the open-source aspect that the project is based on.  The more you tell users what the software can and can’t be used for, the less likely they are to use it.  It becomes more of a burden to follow all of the terms of use than is worth it.  It’s essentially a catch-22.

I think the most obscene part of the whole project is the business structure.  When I buy something, I prefer to be told up-front how much it’s going to cost.  Eucalyptus gives a trial of there software, then you face a fee for continued use.  If you can’t quite figure out the incredibly dense and technical documentation, you’ll be charged a service fee for any help.  If you want to further specialize the software with add-ons, you’ll probably pay for those too.  Welcome to the Americanized version of marketing, where we nickel-and-dime you to death before you even have a working product.

I still think Eucalyptus has a lot of potential, but it requires a great deal of attention to business structure and project management.  Implement release dates, and stop nickel-and-diming to reach profitability.

From the blog jforkey » wsu-cs by jforkey and used with permission of the author. All other rights reserved by the author.

Final Weeks

This week I tried to download a fresh repository and none of my credentials worked. I have been unable to access the new repo to make further changes, so I’m not sure if I will be able to make anymore pushes by the end of the semester. My group has managed to get a decent amount of work done, so I am very excited about that. However, it is unlikely that we will be able to finish the entire project with only a week remaining in the semester and none of us  able to access the new repository.

Overall, it has been fun and interesting learning how to use the Eucalyptus software and writing code and documentation for the eutester package. I wish we could have done more, especially on the development side, but it was a large project and it was difficult to actually get started doing anything. It took the entire semester to get any semblance of a cluster set up, and I still do not believe it is in a working state, not to mention the fact that more than half the class had to learn Python before being able to interpret any of the code. I really wish we were able to develop some real code as a class for the project, or at least finish the project we were working on. There were simply too many variables for an overly ambitious project that contributed to the overall lack of contribution to the Euca community.

Positive things that came out of the semester were that we were able to write a lot of useful documentation for the Euca developers and it seems that all of them are very grateful for our assistance. We met some very smart and interesting people who were enthusiastic about us helping out with their project and I really enjoyed meeting them and talking to them over IRC or email. I wish I was able to meet “mull” when he visited because I had a few questions to ask him, but I had several previous engagements that I could not get out of at the last minute. I think that the code we actually did contribute, combined with the documentation we wrote, will be very helpful to new Eucalyptus developers and testers. I hope that the next group to come along and develop for the project is able to use all of our materials and make some large developments to help out the project. I wish all the Euca people the very best and I hope the project is successful for many years to come.

From the blog trevorhodde by Trevor and used with permission of the author. All other rights reserved by the author.