NoSQL. What it is and how our projects can benefit from it.

An overview of some of the drivers for using NoSQL, the main categories of NoSQL database and what is meant by Polyglot Persistence.

Transcript

1. NoSQL What it is and how Our projects can benefit from it
2. What this is Not ... SQL
3. What it is ...
4. The Plan Not Only SQL Relational Love Polyglot Persistence The Range
5. What is NoSQL?
6. You have a Choice!
7. It’s not just about ...
8. Driving Forces Volume Velocity Variety Cost Agility Mismatch Productivity Availability
9. Volume My application need to be able to handle large volumes of data.
10. Volume Simple Horizontal Scaling Automatic Sharding Distributed Querying
11. Velocity My application needs to store and serve relevant information to users in real-time to provide a high-quality user experience
12. Velocity Distributed Load Store as Used Real-time Analytics Integrated Caching
13. Variety My application needs to be able to handle market data from multiple internal and external systems each with its own schema
14. Variety Handles Sparsely Populated Data Well Dynamic Schema No need to define one schema for all Easier integration of new system’s data
15. Cost We are a public sector organisation which need to account to the local community for our spending.
16. Cost Commodity Hardware Open Source Lower Support Fees
19. Agility My application's data models need to be updated weekly to keep pace with business and market developments.
20. Agility Dynamic Schemas OO Like Structures Easy to Scale Out
21. Productivity My application deals with patient health records and I’m finding it complex to model in my relational database.
22. Productivity Data Complex to Model in Relational Easier to Adopt Avoid Complex Reads/Writes No Upfront Schema Design
23. Mismatch The team developing my application are not database experts
24. Mismatch Store as Used Familiar Formats ORM Redundant
25. Availability My application is critical it needs to be available 24 x 7. I can’t afford a lot of downtime for upgrade.
26. Availability Dynamic Schema Upgrade Without Downtime Simple Replication Simple Load Balancing
27. Key Value
28. Key Value Familiar Simple API Just a BLOB
29. Examples MemcacheDB
30.  The Good Familiar Simple to Build Easy to Scale Great Performance X The Bad XQueries XComplex Data XNo Schema XNot ACID
31. Uses Media Stores User profiles User sessions Shopping carts
32. Document
33. Vs Document { name: “NoSQL”, categories: [“document”, “columnar”, “graph”, “key-value”] }
34. Examples
35.  The Good No Mismatch Productivity Flexible Schema Agile Availability X The Bad X Querying X Transactions X Multiple Uses X Responsibility X Flexible Schema
36. Uses • High Volume Data Feeds • Customer Facing Dashboards • News Sites/Blogs
37. Graph
38. Graph Relations over Data
40. Examples
41.  The Good Relationships Whiteboard Graph Algorithms Schema Free Performance Acid Transactions X The Bad XSpecialised XImmature XPartitioning
42. Uses • Recommendations • Social Networking • Business Intelligence • Fraud Detection
43. Column Orientated
44. Column Orientated BigTable Column, Column Families Ragged
45. Examples
46.  The Good Aggregate Column Fast lookups Distributed storage MapReduce Write Loads X The Bad X API X No Idea of Use
47. Uses Large Deployments Statistics & Analysis
48. Polyglot Persistence
49. Polyglot Persistence
50. Don’t Dismiss Relational
51. Use Relational Transactions Unsure Flexibility
52. Use Relational Integrity Standardisation Tried & Tested Knowledge
53. Use Relational Help Tooling Security & Compliance The Hybrid
54. Questions???