Well, the semester is over now. We gave our presentation in class, and I think it went alright. We’ve got some good information on the wiki now, and more information that isn’t on the wiki right now. We weren’t sure where we should put it, because it didn’t really fit with what we do have. I will be emailing that to the professor.
It would have been nice if we had a working installation earlier to test things out to help us get the information we wanted, but in the end it didn’t work out too bad. In retrospect, it probably would have been a good idea to spend more time in the Eucalyptus IRC channel. Either way, we got done most of what we wanted to get done, although we only managed to get one of the images I wanted done.
I am now actually consolidating the information our group researched and putting it into wiki articles. Right now I have the main page, the cloud controller, cluster controller, walrus and storage controller. I still need to do node controller and I need to go over everything and fix it up a bit.
Our group has also been working on the presentation that we will give during the final exam, and that is coming along well. Link to the wiki below.
Well, the semester is almost over now, and the documentation is starting to shape up. There’s a lot of information that’s hard to track down, though, which is unfortunate. For now, what I am having to do is make an educated guess. Hopefully I will be able to either verify all of the information or someone from the Eucalyptus project or someone who’s had more experience with Eucalyptus can verify or clarify some of the information.
This post is a couple days late because, unfortunately, I’ve been sick. In any case, the documentation is starting to come together, so we should have something useful by the end of the semester. I’ve still been waiting to here about what format the eucalyptus wiki will eventually be, so for the time being we are just keeping everything in plaintext format. Because of the fact that I was sick, I wasn’t able to do a much work as I wanted to, so I am not sure now how many images I’ll be able to produce. Hopefully when we submit our documentation to the eucalyptus project, someone will be able to continue where we leave off.
Now I have been writing the wiki pages for individual components of the cloud. I haven’t actually put them on the wiki yet, I want to flesh them out a bit more first. As I did my research, I realized that there’s a lot I’d like to write about in the wiki, but I am not sure if I will have time to be able to put all of it in before the semester is over. I’ll work on writing what I think are the most important parts, though.
Getting pretty close to being able to write the bulk of the documentation for the overview of the eucalyptus cloud now. Comparing the Eucalyptus user/admin/installing guides to Amazon’s EC2 documentation is helping a lot to close some gaps in information.
I also started working on some images that will help with the documentation. One of them is just a picture to illustrate the general architecture, which isn’t a big deal because such images do exist already. The other is a template that can be used for illustrating how communications with and inside the cloud happen, which I think will be a nice thing to have.
Did some more research this week and found more ways in which Amazon EC2 and Eucalyptus are similar. It looks like amazon’s S3 is pretty close to what Walrus does, and Amazon’s EBS is close to what the storage controller does.
This means that we’ll be able to look at amazon’s documentation to be able to fill in some of the gaps of information in eucalyptus’s documentation, which is good.
Well, we’re into actually working on the project now. I am working on creating some documentation on the general architecture of a Eucalyptus cloud. I feel like this will be a good thing for the project to have. It appears as though the architecture of a Eucalyptus cloud is very similar to that of Amazon’s EC2. That should be a good starting point for describing the Eucalyptus cloud.
Personally, I think that the test we used ( http://www.theopensourceway.org/book/The_Open_Source_Way-How_to_tell_if_a_FLOSS_project_is_doomed_to_FAIL.html ) is pretty silly. The criteria range from mundane to inane, and seem to automatically discount any project that might happen to be windows only (Say, perhaps an open source C#.NET wrapper for DirectX? http://slimdx.org ). I think there is some good information, but other parts of it are strange (Being vaporware means your project is doomed to failure? Who would have guessed?)
In any case, testing eucalyptus with this doesn’t really tell you anything about the project because it’s a long established project.
This week we also split into new groups and started working on seperate projects. I’ll be working on adding documentation to provide a more in depth overview of the eucalyptus architecture.
Well, I am disappointed that I don’t get to work on irrlicht. That project is one that helped me learn C++ and learn about rendering, so it would have been nice to have the chance to give back to the project a little bit.
I have worries about working on eucalyptus, that we won’t actually get to do any software development, that we may end up just trying to fix broken documentation. All I can do at this point is hope that I’m wrong and try to get something done.