Mark: 24 revision(s)

2008-11-03T05:21:47Z

24 revision(s)

New page

[[SE401:Group33|◄ Back]]
==Stuff to look into==

* Existing stress testing/load simulation tools (to be used to test our implementation)
* Scalable software
** What is it
** What are the aims
** What existing infrastructure is there
*** Is it any good?
** Apps currently running on AWS
** Automatically scaling software

* How other people have developed scalable software
** Approaches, techniques, ideas, existing implementations

* Existing tools to automatically scale applications hosted on AWS
** Approaches
** Limitations
** Is the main metric processor utilisation?
*** Is processor utilisation a good metric?

* What are good metrics to determine the need to start/stop application servers to deal with high/low number of requests

==Useful resources==

[http://www.amazon.com/gp/browse.html?node=3435361 AWS Home] 
[http://developer.amazonwebservices.com/connect/forumindex.jspa? AWS Developer Forums] 
[http://www.library.auckland.ac.nz/subjects/eng/resources/eee.htm List of Engineering databases] 
[http://www.library.auckland.ac.nz/databases/learn_database/public.asp?record=ieeexplore IEEE Xplore database] 
[http://www.library.auckland.ac.nz/databases/learn_database/public.asp?record=inspec Inspec database] 
[http://www.library.auckland.ac.nz/databases/learn_database/public.asp?record=acm ACM database] 
[http://www.library.auckland.ac.nz/databases/learn_database/public.asp?record=cia Computer Index Australasia database] 

==Results==

'''Papers'''
*[[SE401:Group33:Research:Mark:Developing Web Services|Developing Web Services]]
*[[SE401:Group33:Research:Mark:TTL Load Balancing|Adaptive TTL schemes for Load Balancing]]
*[[SE401:Group33:Research:Mark:Application Layer Casting|Application Layer Casting - Server Selection of a replicated web service]]
*[[SE401:Group33:Research:Nick:Autonomic_Computing_1| Vision for Autonomic computing]]
*[[SE401:Group33:Research:Nick:Autonomic_Computing_Manifesto| Autonomic computing manifesto]]
 

 
'''Other findings'''
* If an instance fails, all the data stored on it's hard drive is lost, irretrievably. This means a method of persistent storage must be used. Look at http://www.persistentfs.com/documentation/Release_Notes PersistentFS]
* It is currently not possible to create fail safe load balancing in ECS. You cannot guarantee that an instance will run, so you need a dedicated load balancing machine outside the cloud. "Unfortunately, one major drawback with EC2 is that it is not possible to architect a server-side load balancing solution that doesn’t have a single point of failure. Many web applications hosted on EC2 use a single EC2 instance with dynamic DNS to load-balance requests to a particular domain. If the instance that provides the load balancing fails, the whole system can become unavailable until the dynamic DNS maps the domain to another EC2 instance." [http://www.digital-web.com/articles/client_side_load_balancing/ ref]. Therefore look at client side load balancing as describe in ref
* A server side load balancer might be [http://haproxy.1wt.eu/ HAProxy], if server side load balancing was needed
* [http://www.rightscale.com/m/features.html#1 RightScale] do what we want to do<strike>, plus more,</strike> and charge mega mega bucks for it. ''<< Wont be nearly as good as our tool. We will have a more robust system, we wont charge mega bucks... or will we, and our system will be more configurable >>''
**http://wiki.rightscale.com/2._References/Monitoring_System
* http://code.google.com/p/cloudtools/
 

[[SE401:Group33|◄ Back]]

SE401:Group33:Research - Revision history

Mark: 24 revision(s)