Date post: | 29-Dec-2015 |
Category: |
Documents |
Upload: | esther-molly-tyler |
View: | 214 times |
Download: | 0 times |
DMG Confidential 1
DMG STRATEGIC PLANNING
DMG Confidential 2
Background: Key Problem Areas
ScalabilityIngest and export processes not able to handle
burst traffic loadsExponential growth in storage usage and related
costs Development and maintenance
More time now spent on maintenance and support activities versus new development
Dev environment does not support agile application development
DMG Confidential 3
Key Problem Areas continued
Aging code base was not designed to meet current and future demands
Huge demand for DMG services plus focus on short-term benefits led to shortcuts
The technical debt New developers take longer to ramp up New features take longer to develop Testing takes longer Software deployments take longer System has become less stable Extremely difficult to troubleshoot issues
DMG Confidential 4
Background: Scope of Complexity
DMG Confidential 5
Background: Current Plans DMG was looking to leverage MCS to store all DMG
assets and obtain asset management back-end services
DMG would focus on maintaining customer-facing applications
Expected benefits Better scalability and performance Lower long-term storage costs Cost savings from reduced headcounts not needed to support
complex back-end services and corresponding infrastructure
DMG Confidential 6
Background: Assumptions 100% of DMR services would be available in
MCS solution DMR services are functionally and
architecturally adequate MCS can meet new SPE feature requests in a
timely manner MCS charges SPE at cost for custom
development
DMG Confidential 7
Impacts of Assumptions Chose to apply only short-term fixes to
DMR issues rather than implementing complete fixes that would simply end up being “thrown away” once DMR was re-platformed on to MCS
Chose not to work on enhancements if those features are already available in MCS
DMG Confidential 8
Concerns / Risks Will MCS and SPE priorities stay aligned? Less than 50% of DMR services are
enabled in current MCS solution MCS Dev estimates for closing the DMR
and cineSHARE feature gaps are not promising from a timeline or cost perspective
MCS is built on the same legacy code base as EAGL
DMG Confidential 9
Recommendations Find optimal way to leverage MCS services Build a new digital media platform leveraging modern
web technologies, protocols and design practices Leverage open source/best of breed technologies
Ruby on Rails for web developmentElasticSearch for searchMongoDB for high performance data storeNode.js for real-time functionality and mobile APIs
Determine long-term infrastructure strategy
DMG Confidential 10
Approach Migrate users and assets to MCS as required
features become available Leverage MCS for relevant new workflows As a contingency, redirect Acorn retirement
resources to Acorn replacement project based on modern web technologies (new platform)
Assess both MCS and new platform for future potential
DMG Confidential 11
Benefits of New Platform Laser focus on current & future SPE customer needs; no negotiation
necessary Better system monitoring which will lead to improved responsiveness to
customer requests Starting with Zero technical debt
Continuous Integration/Automated Testing Continuous Deployment
AWS delivers ability to handle burst traffic (ingest & export) & storage growth
Smaller development team over time (labor cost savings) Allows for opportunity to revisit certain vendor choices (Civolution, Aspera) Agile approach means that features are delivered quickly as they are built
(rather than waterfall approach of delivering bundles of new features in a single, large deployment)
DMG Confidential 12
New Platform: Architectural Map
Insert diagram