[Esip-cloud] Recap for ESIP Cloud Computing Cluster's telecon on Oct. 24
Huang, Thomas (398G)
Thomas.Huang at jpl.nasa.gov
Fri Oct 28 13:52:30 EDT 2016
Summarizing what we discussed on Monday, I suggest we have back-to-back sessions on Architecting for the Cloud. The first session we invite architects to come share their recommendations and Best Practices. The second session we invite organizations/missions/projects to discuss Cost Analysis. Below are my draft session proposals. I appreciate your inputs and recommendations. I plan to submit our sessions to ESIP later today.
Architecting for the Cloud: Best Practices
Architecting for the Cloud is a non-trivial effort. It is a multi-faceted process. This session invites Cloud architects to come share their experiences and best practices for adapting Cloud solutions in their organizations. Some of the potential areas of interests
* Public, Private, Hybrid Cloud Strategies
* Vendor Evaluation and Selection
* Data Movement and Storage Strategies
* Availability and Redundancy
* Portability vs Proprietary Solutions
* Evolution (incremental) vs Revolution (redesign)
Architecting for the Cloud: Cost Analysis
Commercial and Private clouds have been around for over a decade. Cost is still the key factor inhibiting cloud adoption. While countless studies have been published over the years, estimating the Total Cost of Ownership is still a domain-specific effort. This session invites presentations focusing on cost analysis conducted by organizations, missions, and projects. Some of the potential areas of interests
* Instances Types Selection
* Speed vs Cost
* Long-term Storage and Availability
* Cost Saving in Proprietary Solution
* Tools and/or Process to estimate Total Cost of Ownership
Jet Propulsion Laboratory
4800 Oak Grove Drive, Mail Stop 158-242, Pasadena, CA 91109
Phone: 818.354.2747, Email: thomas.huang at jpl.nasa.gov<mailto:thomas.huang at jpl.nasa.gov>
DISCLAIMER: All personal and professional opinions presented herein are my own and do not, in any way, represent the opinion or policy of JPL, NASA or Caltech.
On Oct 24, 2016, at 12:24 PM, Fei Hu via Esip-cloud <esip-cloud at lists.esipfed.org<mailto:esip-cloud at lists.esipfed.org>> wrote:
Thanks for attending today’s telecon. The recap for the meeting is as follows. If you have any revision, please update it at the wiki pagehttp://wiki.esipfed.org/index.php/Cloud_Telecons_10/24/2016.
* How did NASA use commercial cloud, such as Amazon? What kind of projects need to set up its own cloud in an affordable manner. One of Amazon advantages is to scale up resources. The private could need consider the largest resource for the worst case. On Amazon, you just need pay as you go.
* When to make sense on cloud or hardware. Looking into the provision service, such as mapreduce, nosql, to accelerate their system
* Technical session: how do people apply cloud computing on their scientific researches.
* The comparison between Google computing engine, Microsoft Azure, Amazon, Redhat, etc. Look into their features, differences, and potential cost. The cost estimator from GMU for different cloud platforms can be useful, but the information need be updated. Two keys for cloud service users: 1) how much is the cost overtime? 2)how to grow the system overtime?
* GPU to accelerate the analysis/software on Cloud; how much the cost of GPU on Amazon? Several issues with GPU may need to be solved: data into GPU/ data movement/ IO bottleneck.
Summary for session proposal:
* comparing of using public cloud and private cloud.
* comparison between different venders: google redhat etc.
* some researches that move to cloud and truly leverage the provision service
* application of cloud (including big data) to solve their projects or missions
* send out the session proposals by this week
Esip-cloud mailing list
Esip-cloud at lists.esipfed.org<mailto:Esip-cloud at lists.esipfed.org>
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Esip-cloud