Apache Cassandra Lunch #103: Cassandra Cluster Architecture in UML and the Azure Digital Twin Domain Language
Download as PPTX, PDF0 likes55 views
The document discusses the architecture of a Cassandra cluster using UML and the Azure Digital Twins Domain Language (DTDL). It outlines the three layers of architecture (functional, generic physical, and specific physical) and their respective outcomes, as well as major operations and services involved in the cluster. Additionally, it highlights the capabilities and limitations of Azure Digital Twins Explorer for modeling and analyzing digital twins.
Apache Cassandra Lunch #103: Cassandra Cluster Architecture in UML and the Azure Digital Twin Domain Language
- 1. Version 1.0 Architecture of a Cassandra Cluster in UML and the Azure Digital Twin Domain Language UML Architecture and Exploring Azure Digital Twin Explorer and DTDL for Cassandra Nicholas Brackley Enterprise Architect
- 2. What is Architecture? ● “the conceptual structure and logical organization of a computer or computer-based system.” ● “Makeup of a composite within context” ● That building thing
- 3. 3 Layers of Architecture and their Outcomes ● Functional Architecture ○ Data that needs to be managed ○ Database definitions ● Generic Physical Architecture ○ Service and Component Definitions ● Specific Physical Architecture ○ Service Selection, Deployment Architecture, and Digital Twin
- 4. Functional Architecture of a Cluster ● What are the concepts/ideas involved in the Cluster ○ Ingress and Egress ○ ELT ● What are major operations/engagements occuring? ● Nouns, verbs, and adjectives in UML ○ All terms in Cassandra.yaml and other config files are captured in the functional architecture
- 5. Generic Physical Architecture of a Cluster ● Answers Questions ○ What are major services and where are they located? ○ What are the structural divisions in your system? ● Uses ○ Training ○ Deployment Agnostic Architecture ■ Transition Between Clouds/Services ■ Service Connection Requirements
- 6. Sample Generic Physical Architecture
- 7. Specific Physical Architecture of a Cluster ● Service Selections ● What services are being deployed into what infrastructure?
- 8. Cassandra Cluster Specific Physical Architecture
- 9. Cassandra Cluster Specific Physical Architecture https://github.com/Anant/cassandra.toolkit
- 10. Architecture of a Cassandra Cluster https://blog.anant.us/apache-cassandra-lunch-51-cassandra-cluster-design-and-architecture/ *semi-Functional *semi-Physical
- 11. Digital Twins, What and Why? ● A Simulated System ○ Modeling for testing ○ Life-cycle analysis ○ Cost savings ● Use Data Inputs from Real Systems Validate Models ○ Compare the inputs and outputs of the real system and Digital Twin ○ Tools to understand zero day failures ● Tools for Training and Analysis
- 12. Azure Digital Twins Explorer ● Released June 1, 2021 – Available in Preview on browser in Azure Marketplace ● Import models in human readable formats (JSON, xlsx) ○ Note that this primarily uses Digital Twin Domain Language (DTDL) which is based on JSON- LD ● View Relationships (inheritance, composition) in a graph view ● CLI with SQL-like language to throughout model ● Properties editable ● Connect Properties to IoT devices in the real world
- 13. Limits to DTDL in Azure ● All top-level DTDL elements in a model must be of type Interface. The reason for this requirement is that Azure Digital Twins model APIs can receive JSON objects that represent either an interface or an array of interfaces. As a result, no other DTDL element types are allowed at the top level. ● DTDL for Azure Digital Twins must not define any commands. ● Azure Digital Twins only allows a single level of component nesting, meaning that an interface that's being used as a component can't have any components itself. ● Interfaces can't be defined inline within other DTDL interfaces; they must be defined as separate top-level entities with their own IDs. Then, when another interface wants to include that interface as a component or through inheritance, it can reference its ID. - There is a LOT of editing necessary to propagate changes throughout a model.
- 14. Demo (stage 1 - establishing models) ● Using Azure to Start the Instance ● The Room and Floor Tutorial ○ Query ● Static Cassandra Cluster Example ○ Relationships, Inheritance, and Components ○ Query
- 15. Strategy: Scalable Fast Data Architecture: Cassandra, Spark, Kafka Engineering: Node, Python, JVM,CLR Operations: Cloud, Container Rescue: Downtime!! I need help. www.anant.us | solutions@anant.us | (855) 262-6826 3 Washington Circle, NW | Suite 301 | Washington, DC 20037