1 / 24

Feedback from Summer Camp Students

Feedback from Summer Camp Students. July 2013 www.geni.net. GREE-SC13 Summary. June 24 th – June 28 th @ University of Connecticut Wireless theme 7 Tutorials on multiple GENI tools covering the GENI Experiment Life Cycle. Summer Campers.

jane
Download Presentation

Feedback from Summer Camp Students

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Feedback from Summer Camp Students July 2013 www.geni.net

  2. GREE-SC13 Summary • June 24th – June 28th @ University of Connecticut • Wireless theme • 7 Tutorials on multiple GENI tools covering the GENI Experiment Life Cycle

  3. Summer Campers • 17 attendees (14 registered + 3 local students) (ugrad, grad, professor) • 5 Final Projects w/ live demos! 2 Demos at GEC17 based on SC projects

  4. Projects • Wimax, surveillance network NYU Poly + ORBIT testbeds • VM Migration ProtoGENI with OVS + GIMI • Path restoration ProtoGENI w/ OVS + GEMINI • User-controlled flows-policies ExoGENI w/ OV + GIMI • Resilient routing ExoGENI w/ OVS + GIMI (switched to ProtoGENI)

  5. 1. Adaptive Flow Control Using OF • Diverse application bandwidth requirements • Over provisioning or under provisioning of network resources • Tight coupling of routing tables on switches does not allow the user to control the flow policies Motivation: Rate control the bandwidth based on application flows instead of arbitrary allocation of resources

  6. 1. Topology ExoGENI+GIMI Host 1 Host 3 OVS 2 OVS 1 Host 2 Host 4

  7. 2. Resilient Networks with DAG • Link failures cause packets loss. • Goals • Demonstrate understanding of GENI terminology, tools, and resources. • Reduce packet loss. • Boot-strapping (create DAG) • Generate a random directed acyclic graph (DAG) • Make the random DAG as destination oriented DAG \

  8. 2. GENI Resources • Aggregates: Utah ProtoGENI • OpenFlow Switches • Five raw PCs running OpenvSwitch • OpenFlow Controller • Raw PC running POX

  9. 2. GENI Resources

  10. 3. VM Migration • Live migrations are expensive • Additional network traffic generated. Affects performance of other application in the network • Downtime of applications on migrating VM • Vendors recommend a separate network for VM mobility – costly

  11. 3. Network Topology • All resources reserved from ProtoGENI AM • 2 Raw pcs running ovs • 3 VMs • GEMINI extensions

  12. 4. Network Path restoration using OpenFlow

  13. 4. Topology ProtoGENI + GEMINI

  14. 5. WiMax Mobile Surveillance for Law Enforcement • High quality video & audio over fast-moving vehicular environments • Secure data streaming • Real-time command and control • Efficient group management

  15. 5. NYUPoly Witest

  16. Questions to Summer Camp Attendees • What is your research about? • Which parts of GENI would be most useful for your research? • What in GENI worked well at the camp? • What is missing from GENI in terms of tools, resources, anything that will make your research-life easier? Got responses from 9 out of 17 attendees

  17. 1. What is your Research About? • WiMax, opportunistic networks • Cloud storage systems • VM migration and Netflow data analysis • Large scale failure analysis and recovery in backbone networks; network situational awareness systems • Path diversity and geodiversity for routing protocols

  18. 1. What is your Research About? • Security and Privacy • Vehicular network • Social Networking + Distributed System • Performance evaluation of VM Migration using OpenFlow A lot of interest in OpenFlow

  19. 2. Which parts of GENI would be most useful to you? • WiMax Orbit, Witest • OpenFlow • Large scale testing • Ability to test new routing protocols

  20. What Worked Well? • Tool diversity • The tutorials are very helpful, GENI people are very knowledgeable and helpful! (multiple comments) • Every part of the summer camp

  21. What Did not Work Well? • I was still not getting to used to use all the resources because of the time limitation • The network and power need some care • I failed to follow the WiMax part because of the keys issue • The delay and instability of resources.

  22. What is Missing from GENI? • More stability • Tools for post-processing the experiment the data, for example, packet delivery ratio, delay • More sample codes / case study • Mobile devices as resources to communicate with each other and track their mobility

  23. What is Missing from GENI? • My works deals with processing big data. That would be really helpful if GENI supports planned data uploading / storing (for a significant period of time) • If all resources are used up, can we have a way of knowing that they are. And depending on experiment allocation time, the expected duration after which some resources may be released. • If we have resources already ready in our slivers, and we change one node, can we reallocate that node while the rest of the nodes get not affected and still continue to function • I.e. they want Update()

  24. Campus Network Issues • Campus IT blocked traffic to “non-standard” ports • GENI uses a number of “non-standard” ports • We could not give IT a list of ports to open • Not documented • Different networking problems on different campuses • E.g. Rochester blocked traffic to port “12356” but allowed traffic to other ports • Even if ports are open, firewalls block access when they see large numbers of connections to the same port

More Related