Download as PDF, PPTX
Uploaded byOtávio Santana
NoSQL design pitfalls with Java
The document discusses the pitfalls of NoSQL design with a focus on Java applications, highlighting the misconceptions and challenges associated with systems like Apache Cassandra and MongoDB. It emphasizes the need for proper modeling, the importance of consistency levels, and the complexities of cloud deployment. Platform.sh is presented as a PaaS solution that simplifies cloud application development and management, promoting a polyglot approach and continuous deployment.
More Related Content
Similar to NoSQL design pitfalls with Java
![Architecting Cloud Computing Solutions with Java [1.1]](https://cdn.slidesharecdn.com/ss_thumbnails/architectingcloudcomputingsolutionswithjava-200728160449-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
![Jakarta EE Meets NoSQL in the Cloud Age [DEV6109]](https://cdn.slidesharecdn.com/ss_thumbnails/jakarta-nosql-cloud-181025142320-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
![Let’s Make Graph Databases Fun Again with Java [DEV6043]](https://cdn.slidesharecdn.com/ss_thumbnails/graph-181025142147-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
![Eclipse JNoSQL: One API to Many NoSQL Databases - BYOL [HOL5998]](https://cdn.slidesharecdn.com/ss_thumbnails/jnosqlhol2018-181025141725-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
NoSQL design pitfalls with Java
- 1.
- 2. Otavio Santana @otaviojava DevRel Engineer +Java Champion + JCP-EC-EG-EGL + Apache Committer + Eclipse Committer + Eclipse Project Leader + Book and blog writer Speaker
- 3. What it isnot!
- 4. ● “New researchshows 75% of ‘open’ Redis servers infected” ● The Twelve Factor App NoSecurity
- 5. ● Schemaless ● Validation NoData integrity
- 6. ● Complexity ● Maintain NoResponsibility
- 7. ● The rightquestions ● Horizontal vs Vertical ● Write vs Read Scalability??
- 8. ● Benchmark ● CAPtheorem No-mistake
- 9. ● CAP theorem ●Backup No-Trade-off
- 10. ● SQL ● NewSQL ●NoSQL No-single-solution
- 11.
- 12. ● Key ● Biggestmistakes ● View Modeling
- 14. What Apache Cassandrais not
- 15. ● Big Data ●Rich Query ● Relational Cassandra
- 16. ● Key ○ PartitionKey ○ Clustering Key ● Secondary Key ● Allow Filtering Queries
- 17. ● List ● Map ●Tuple ● UDT Cassandra Powers
- 18. ● Load Balancer ●Consistency Level ALL ● Backup Troy
- 19. Sample @Entity public class Contact{ @Id private String name; @Column private LocalDate birthday; @Column private Map<String, String> details; } ● Map ● UDT
- 20. Sample @Entity class Car{ @Id private String plate; @Column private String city; @Column private String color; @UDT("owner") @Column private Owner owner; } ● UDT
- 21. Sample @Entity public class Recipe{ @Id private String name; @Column private String city; @Column @UDT("ingredient") private Set<Ingredient> ingredients; } ● Set ● UDT
- 22.
- 23. ● Transaction ● ACID WhatMongoDB is not
- 24. “In a relationaldatabase, we usually model the schema at the very beginning of the project. Why is this so bad? Because the beginning of the project is exactly the moment that we know least about it” Oren Eini - Modeling in a Non Relational World
- 25. ● Coherence ● Independence ●Isolation DDD - Elemar Jr
- 26. ● Self Sufficient ●Business Coherence
- 27. ● Exist ● Itdoes not need anyone Independence La liberté guidant le people (1830)
- 28. ● Cannot impactanother entity Isolation
- 29. ● Entity ● Aggregate ●Value Object Sample @Entity public class Order { @Id private ObjectId id; @Column private LocalDateTime orderedAt; @Column private List<Product> items; @Column private Address shipTo; }
- 30.
- 31. ● Containers ● Mutable ●Kubernetes ● Services Cloud
- 32. ● Bean Validation ●ORM ● Reactive programming ● Impedance Java Frameworks
- 33.
- 34.
- 35. We live ina cloud economy Use cloud ready SaaS services Complexity Cost Risk High Med - High Low Migrate existing workloads into cloud Build new apps cloud ready Degrees of risk around enterprise cloud consumption... ^ ^
- 36. We’re a polyglot,multicloud PaaS, with continuous deployment built in. At Platform.sh we believe your code should just run. Monoliths? Microservices? Stateful? Stateless? Develop and deliver them all with consistent tools. Much more than hosting
- 37. Infrastructure as code Add MySQL,PostgreSQL, MongoDB, Kafka, Elasticsearch, Solr, Redis, RabbitMQ, or InfluxDB to your project with a few lines of code. Infrastructure changes are versioned and auditable. Service updates and security? That’s on us. No more effort wasted on patching your fleet.
- 38. Application name: app type: "java:11" disk:1024 hooks: build: mvn clean package relationships: database: "db:postgresql" web: commands: start: java -jar $JAVA_OPTS $CREDENTIAL file.jar
- 39. Services database1: type: mysql:10.1 disk: 2048 database2: type:postgresql:9.6 disk: 1024
- 40. Route "https://{default}/": type: upstream upstream: "app:http" "https://www.{default}/": type:redirect to: "https://{default}/"
- 41. Platform.sh provides an end-to-endPaaS to build, run, and maintain polyglot fleets of digital experience applications Developers can then focus on what matters to the organization: shipping great web apps and sites Container Orchestration Infrastructure Continuous Deployment Data Services App Runtimes Routing + Edge Security Continuous Integration Application Code Application Data Provisioning APIs, Integrations, Template Libraries Web Console … Source Operations™
- 42. 1000s e-commerce, life sciences,government, education, media & entertainment, and high-tech customers $47m invested by top-tier international partners Platform.sh at a glance
- 43. Q&A Thank you! OtavioSantana DevRel Engineer, Platform.sh @otaviojava