SERVICE

Journals
May 4, 2009

Gabriel
  • Met with Dr. Feng, Billy, and two new students that wish to work with us in the coming year: Michelle and Scott.
  • We spoke of the work to be done next semester. For Billy and I this includes working more with Storytelling Alice, including porting it to Linux so that we can run it in our VM instances.

Billy
  • Attended the meeting with Dr. Feng, Gabriel, and the new undergraduate researchers, Michelle and Scott.
  • Looking forward to working with new people and taking the next step in our ultimate goal of providing a service to underfunded school systems.
May 1, 2009

Gabriel
  • Met with Billy and Mark to finalize the direction for our research in the coming year.
  • Suggested that Billy create a CD with the scripts and programs to be run.

Billy
  • Discussed the directions of our research for next year.
  • Found evidence that NAT does not work in Xen HVM currently.
  • Started creating an iso with the necessary test files.
April 24, 2009

Gabriel
  • Continued helping Billy set up a Windows VM in Xen.
  • Running into issues with getting necessary scripts into the VM.

Billy
  • Continued working on the networking issues with Xen.
  • Explored other ways of getting the files into the VM.
  • Mounting the hard drive image did not work due to the format of the image.
April 17, 2009

Gabriel
  • This week we worked with students at Harding Avenue Elementary School, once more with Storytelling Alice.
  • Continued analyzing data.

Billy
  • Setup a Windows VM in Xen.
  • Need to work on overcoming some networking issues with Xen to run tests on the Windows VM.
  • Participated in teaching Storytelling Alice at Harding Avenue Elementary.
April 10, 2009

Gabriel
  • Once again worked with students at Blacksburg New School on their stories in Storytelling Alice.
  • Completed draft of write-up.
  • We have begun analyzing the large amount of information we got from the tests.

Billy
  • Finished the testing on Windows VMs.
  • Ran scripts to make the data into graphs to see the trends more clearly.
  • Put the data from the tests into the write-up.
  • Discussed possible ways to work with Storytelling Alice on virtual machines.
  • Participated in teaching Storytelling Alice at Blacksburg New School.
April 3, 2009

Gabriel
  • Helped students learn and use Storytelling Alice at Blacksburg New School.
  • Figured out the issues with Xen and finally ran tests.
  • Finished all network performance tests.
  • For the next week, my goal is to write up what I have done so that it can be reproduced.

Billy
  • Setup rdesktop for the Windows VMs to simplify access.
  • Installed perl on Windows VMs and ran basic network tests.
  • Modified more intensive perl testing scripts to be executable on Windows.
  • Started running more rigorous networking tests.
  • Helping Gabriel to make the write up.
March 27, 2009

Gabriel
  • Still having trouble booting a VM in Xen.
  • Using hardware virtualization, I was able to boot a VM, will be testing this in addition to the paravirtualized VMs.
  • For this week my goal is to finish up network performance tests.
  • Also, I am to begin a write-up of all the work I have done.

Billy
  • Worked on setting up two network cards on the Windows VMs for simplified use.
  • Began copying necessary testing scripts to Windows VMs.
March 20, 2009

Gabriel
  • Tested VirtualBox using hardware virtualization support.
  • Ran tests in KVM utilizing paravirtualized network driver.
  • Problems booting Xen VM, will continue working on it for next week.
  • Began playing with hardware virtualization in Xen.

Billy
  • Finished loading iperf on Windows VMs.
  • Testing Storytelling Alice on WINE. Unable to get graphics to show up.
  • Working with some network issues in the Windows VMs. Need to directly assign ip addresses.
  • Begin network tests of Windows VMs.
March 13, 2009

Gabriel
  • Spring break this week

Billy
  • Spring break
March 6, 2009

Gabriel
  • Finally got bridged networking in KVM.
  • Ran network tests in VMware and KVM.
  • Will need to finish setting up Xen and running the initial tests.

Billy
  • Finished installing XP to all VMs.
  • Need to download and install iPerf on Windows VMs.
  • Need to work out some networking issues in order to run tests.
February 27, 2009

Gabriel
  • Ran network tests on VirtualBox.
  • Finally figured out how to access VMware's web-based interface, though some issues still remain.
  • Still working on setting up a bridged network for KVM.

Billy
  • Installed XP on Virtual Box VM and created an image to install on for KVM.
  • Will work with Gabriel on installing XP on VMware due to some complications with new web-based interface.
  • Continue looking into playing games in Linux through Wine.
February 20, 2009

Gabriel
  • Week spent sick, but got work done
  • I have completed my custom LiveCD. Now testing will be very streamlined.
  • In searching for a Xen-enabled kernel, I have found two different answers. I will be trying each and deciding upon the best solution.
  • For this week I will continue testing and help Billy in installing Windows XP.

Billy
  • Also sick for a week.
  • Install Windows XP on the Woodcrest machine VMs.
  • Continue looking into playing games in Linux through Wine.
February 13, 2009

Gabriel
  • Work on the custom LiveCD and on the Xen kernel continues.
  • I got the first tests running on the bare hardware.

Billy
  • Discussed progress with Gabriel.
February 6, 2009

Gabriel
  • As Ubuntu has dropped official support for Xen, I will need to get a kernel capable of working with Xen this week.
  • To streamline the testing process, I will also be creating a custom Ubuntu LiveCD.

Billy
  • Attempt to run some educational games on WINE.
  • Continue searching for educational games for both Windows and Linux
January 30, 2009

Gabriel
  • Back for a new semester, continuing where I left off.
  • Run the network tests on the Woodcrest machines now that thye are set up with new versions of the VM software we were using.
  • Help Billy get up to speed so that we can move on with the testing in parallel.
  • Keep doing literature searches in the background.

Billy
  • Look at the WINE program for linux.
  • Search for and check out new educational games.
  • Help run new performance tests on Woodcrest machines.
  • Continue searching for useful literature.
December 5, 2008

Gabriel
  • At the talk I gave on my research, one person in the audience asked a good question about Xen's performance. As such, I will be testing Xen further, adding even more processes to see how its performance degrades.
  • For the next week we will be setting up two new machines to test on which have hardware support for virtualization. We will then be re-running some tests so as to have a base to compare against.

Billy
  • Look into using Immune Attack for pedagogy.
  • Search for research papers that may be useful for reference.
  • Attempt to gain access to and set up new VM host machines.
November 22, 2008

Gabriel
  • I have returned from SC|08 with a poster presentation, a talk, and much work behind me. Time for a little rest...

Billy
  • Enjoy Thanksgiving break while it lasts.
November 14, 2008

Gabriel
  • I will be at SC|08 with Dr. Gardner, Dr. Feng, and two graduate students from Tech.
  • Have to prepare to talk about my poster and give a presentation on my research thus far. The poster will be uploaded as soon as possible.

Billy
  • Continue gathering profiling data.
  • Wrote a perl script to store profiling data from the Kernrate program in a more useful format.
November 7, 2008

Gabriel
  • I am getting all my data together and finishing the poster this upcoming week.
  • Also, I will be preparing for my upcoming trip to Austin, Texas.

Billy
  • Continue gathering profiling data.
Octorber 31, 2008

Gabriel
  • For the next week I am to focus on my poster for Supercomputing '08, which I will present in mid-November.

Billy
  • Begin gathering profiling data using Kernrate.
Octorber 24, 2008

Gabriel
  • Dr. Gardner, Billy, and I spoke at length about the direction the project would eventually be taking, in regards to our work in using virtualization for pedagogy.
  • I continued gathering information for my poster.

Billy
  • Get iperf script for testing from Gabriel.
  • Search for a Windows version of iperf.
  • Attempt to get basic profiling data.
  • Make sure NAT works in Qemu.
Octorber 17, 2008

Gabriel
  • Met with Billy, Dr. Feng, and Dr. Gardner this week. Billy and I reported the progess we had been making. Following this we discussed the work we would need to accomplish to meet our goals. For me this was mostly my profiling work shaping my results to fit a poster for SC|08.
  • Dr. Gardner spoke at length with Billy and I on the topic of networking. The main reason was to help us understand the differences between NAT, host-only, and bridged networks in virtual machines.

Billy
  • Download and install Qemu
  • Begin setting up for profiling in Windows
  • Ensure NAT works in VirtualBox VM
October 15, 2008

Gabriel
  • Had a meeting with Billy to finalize some sections on the SERViCE web site.
  • Still a bit more work to do before we make the changes public.

Billy
  • Work on finalizing the website.
  • Research Oprofile.
October 3, 2008

Gabriel
  • Regenerated profiling charts for QEMU, VirtualBox, and VMware Server. Used Oprofile again but made sure to record more details.
  • Continued working on editing the iperf source code. I want to get an accuracy of nanoseconds for the timestamps. As a result, such I have been trying to use the CPU counter to calculate them.

Billy
  • Download and install VirtualBox.
  • Create a VM in VMware with a bridged connection.
  • Create a VM in VirtualBox with a bridged connection.
September 25, 2008

Gabriel
  • Spoke with Dr. Gardner and Dr. Feng about the requirements for the web site along with some ideas. In the end we decided to create a journals page, a people page, and possibly a page for higher-level news.
  • Dr. Gardner and I further discussed my work on iperf. He suggested I look up a "read timestamp counter" function. In addition, we discussed the expectations we would have in the differences between running iperf natively and inside of a virtualized operating system.

Billy
  • Research more information about virtualization.
  • Continue to work on getting the website up and running as desired.
September 16, 2008

Gabriel
  • Talked with Dr. Gardner for a while about my the goal for my edits in the iperf source code. He explained that I should print the timestamp just before a packet is sent and just after the confirmation is received. We sat down and discussed how the packet would travel from the guest OS in the virtual machine down to the host OS and out to the network.

Billy
  • Get a better understanding of optimizing memory storage in a VM through defragmentation.
  • Work on website.
September 12, 2008

Gabriel
  • Spoke with Dr. Gardner and showed him the profiling results. He recommended that I use the information to try to draw up a call graph.
  • Also decided that, in order to get a finer-grained view of a network packet's path, I should recompile iperf. Up to this point I had been using it to test network throughput, but now require more information than just bandwidth.
  • Need to bring a Windows XP disk for Billy for him to properly install VMware Server.

Billy
  • Set up a VM with internet access using NAT.
  • Reinstall VMWare after installing Microsoft Internet Information Services.
September 5, 2008

Gabriel
  • Completed 10 minute presentation to give to the SERViCE group, though it looks like it will be pushed back.
  • Still working on profiling.
  • Gave quick 3 minute presentation in Synergy meeting giving a brief overview of my work.
  • Met with Dr. Gardner, Dr. Feng, and Billy and discussed the project in more detail, including networking and security.

Billy
  • Download and install VMWare.
  • Install Ubuntu and become familiar with the VM.
August 29, 2008

Gabriel
  • First Synergy group meeting this week.
  • Met up with Dr. Feng and Dr. Gardner for the first time this semester. Covered expectations for research and also what to work on the following week.
  • Decided to profile QEMU and VirtualBox and also need to prepare presentation for next week.

Billy
  • Move in.
  • Discuss proposed weekly meeting time for research.