Delphix has enjoyed great success and acclaim for our capabilities around databases, but not many people know that we can deliver the same benefits for applications via our "unstructured data" dataset type. I documented my experience using Delphix to virtualize an ERP application (BMC Remedy) on my personal blog: http://http://therealcloudsurgeon.com
. Others, like Piotr, have also documented some of their experience with other applications ( http://kolibero.blogspot.com/
). The possibilities are virtually endless.
What is important. is to understand is the framework for virtualizing new applications. The good news is that companies and organizations that are used to automating applications will be able to easily virtualize the applications, too. Below, I outline my basic thought process (non-exhaustive) when approaching a new application with Delphix. It is important to note, the answers these questions each beg three more questions: "Why?", "Is it automated?", and "If not, can it be automated? (Why Not?)" Automation is not a requirement to gain the benefits of virtualization, but automation usually a key indicator that the application is a good candidate for virtualization.
Understand the Architecture of the Application
- How many tiers?
- What is the purpose of each tier?
- How does each tier communicate with each other?
- How many nodes?
- Does the # of nodes and tiers change? (i.e. Dev vs QA vs Prod)
- Is the application stateless?
- Does the application leverage/utilize an RDBMS?
- Does the application leverage the file system for caching or high-transaction temp?
- After the initial installation and configuration, what changes in the environment? (i.e forms, customizations, etc)
- Where do those changes get written? (i.e. certain directories on the file system, the database, etc)
- What is the I/O driven at the various components (filesystem, dba, etc) ?
Understand the Backup, Restore and Continuity of Operations Process of the Application
- How is the application backed up for DR?
- How is the application configured for HA?
- If the power plug on one, or all, of the application tiers were pulled, what would be the process to get the application running again?
Understand the Non-Production Lifecycle
- How are non-production application instances created?
- How are non-production application instances refreshed?
- How are non-production application instances reset after destructive actions? ( i.e. destructive testing, misconfiguration, corruption"
- If non-production instances are masked, how is the masking performed today?
- What I/O loads are typical for non-production instances?
I will post again soon about where to go from here. Leave your comments and feedback. Did I miss something? I will do my best to address any questions posted.#DemoEnvironment