Understanding application requirements
0 likes255 views
The document discusses cloud computing and how to properly size cloud resources for applications. It defines cloud computing as on-demand access to configurable computing resources over a network. Key aspects are ubiquitous access, rapid provisioning and release of resources with minimal management, and sharing resources in a configurable pool. Properly sizing resources involves understanding both the supply of offerings from cloud providers as well as the demand requirements of the specific application. Factors like CPU, storage, memory, and network needs must be assessed based on the application's usage patterns to ensure optimal performance.
Understanding application requirements
- 2. Evolution (not a revolution) 2 ¨ Cloud Computing is just computing but what exactly is cloud computing? n If you ask 100 people to define cloud, you may possibly get 100 different answers ¨ Formal definition from the National Institute of Standards and Technology: ¤ Page 3 of the final pub: http://csrc.nist.gov/publications/nistpubs/800-145/SP800-145.pdf Cloud computing is a model for enabling ubiquitous, convenient, on-‐demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction. ¨ Cloud computing is more about evolution than revolution, with technology allowing price points and make them available to all people — regardless of budget size — via a utility-based, pay-per-use model.
- 3. Ubiquitous, convenient, on-demand access 3
- 4. Rapidly provisioned and released 4
- 5. Minimal management effort 5 ¨ Negligible interaction with the service provider
- 6. Shared pool of configurable resources 6 ¨ Right-sizing is key Demand Supply Resource Resource Resource Resource Resource Shared pool of resources: • Networks • Servers • Storage • Apps, and • Services End users suffer Cost shoots up
- 7. Understanding the supply side 7 ¨ Carefully understand cost implications ¨ Evaluate suppliers’ offering ¨ Geographic location/proximity to end users ¨ OS/Arch type and license implications ¨ Machine sizes ¨ Network bandwidth and storage costing
- 8. Lets examine what sizes Amazon offers 8 US East: Virginia Pricing as of late 2011
- 9. And what sizes Rackspace offers 9 US Pricing as of late 2011
- 10. Providers make it easy to calc costs 10 ¨ Rackspace Cloud Calculator ¤ Try this Calculator in class ¤ www.rackspace.com/cloud/ cloud_hosting_products/servers/pricing ¨ Example: ¤ Cost = 105.60 monthly to run a Rackspace server with 2GB memory running Linux (except Red Hat)
- 11. Understanding the demand side 11 ¨ Developers typically call out system requirements ¤ http://drupal.org/requirements ¤ http://wordpress.org/about/requirements ¤ http://www.joomla.org/technical-requirements.html ¤ https://github.com/instructure/canvas-lms/wiki/Production-Start ¨ But are sometimes silent on size requirements ¨ Size requirements really depend on usage
- 12. Assessing the right size ¨ CPU ¤ size(s) and count ¤ Geo location ¨ Storage size ¤ Block storage ¤ File storage ¤ Database storage ¨ Memory size ¤ Memory the app needs ¨ Network bandwidth ¤ among server nodes ¤ between end user clients and server nodes in the cloud 12 Size estimates depend on usage patterns
- 13. App Performance Profile 13 ¨ Application may be ¤ CPU intensive, ¤ memory intensive, ¤ disk intensive, ¤ network intensive, or ¤ a combination of all of the above.
- 14. Questions and discussion on topic 14
- 15. 15 Thank you.
- 16. Storage sizing 16 ¨ Important data that can help you make storage architecture decisions ¤ Storage allocated ¤ Storage used ¤ the disk I/O profile of the server ¤ disk queue length ¤ read versus writes ¤ paging ¤ swapping
- 17. Logical grouping your cloud instances 17 ¨ If you group applications or VMs together that have the same recovery requirements, they can more easily replicate storage.