O slideshow foi denunciado.
Seu SlideShare está sendo baixado. ×

Using Cloud Automation Technologies to Deliver an Enterprise Data Fabric

Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
Anúncio

Confira estes a seguir

1 de 37 Anúncio

Using Cloud Automation Technologies to Deliver an Enterprise Data Fabric

Baixar para ler offline

The world of database management is changing. Cloud adoption is accelerating, offering a path for companies to increase their database capabilities while keeping costs in line. To help IT decision-makers survive and thrive in the cloud era, DBTA hosted this special roundtable webinar.

The world of database management is changing. Cloud adoption is accelerating, offering a path for companies to increase their database capabilities while keeping costs in line. To help IT decision-makers survive and thrive in the cloud era, DBTA hosted this special roundtable webinar.

Anúncio
Anúncio

Mais Conteúdo rRelacionado

Diapositivos para si (20)

Semelhante a Using Cloud Automation Technologies to Deliver an Enterprise Data Fabric (20)

Anúncio

Mais recentes (20)

Anúncio

Using Cloud Automation Technologies to Deliver an Enterprise Data Fabric

  1. 1. Using Cloud Automation Technologies to Deliver an Enterprise Data Fabric Sean Martin, CTO Cambridge Semantics Inc. DBTA, 21 April 2020
  2. 2. • Big data volume • Ad hoc queries • Unstructured • Semi-structured • Exploratory • Raw • Self service • On-demand • Clean • Consistent • Integrated • Accessible • Searchable • Secure • Governed • Privacy (PII) • Clean • Consistent • Integrated • Accessible • Searchable • Secure • Governed • Privacy (PII) • Clean • Consistent • Integrated • Accessible • Searchable • Secure • Governed • Privacy (PII) • Big data volume • Ad hoc queries • Unstructured • Semi-structured • Exploratory • Raw • Self service • On-demand • Cataloged • Linked • Modeled • Persisted • Virtualized • Collaborative Data Fabrics are the modern successor to warehouses and lakes. Fully connected and integrated Data Fabric Data Lake Data Warehouse Flexibility and scale Quality and control
  3. 3. RDBMS/OLTP Big Data / Hadoop Document Repositories Traditional BI Cloud CLAIM CUSTOMER PRODUCTS POLICY Semantics and graph allow the data fabric to be an overlay spanning and encompassing the existing data and analytics landscape.
  4. 4. • Patients • Encounters • Providers • Medications • Costs • Care Plans • Claims • Etc. Providers Care Plans Patients Costs Inpatient Claims Carrier Claims Outpatient Claims Prescriptiom Drug_Events Beneficiary Summary BestPractiseLinks careprog2 careprog1 Medications Patient Encounters Observations Conditions Allergies Patients Procedures Imaging Studies Immunizations Care Plans care planscanonicalelectronic medical records claims How it works: Business Friendly Data models Semantic Graph data models to capture and navigate data relationships
  5. 5. Real World Graphs Get Big Fast Vast Hundreds of sources, representing thousands of entity types Siloed Different technologies, schemas, formats Complex Sprawling disconnected schemas, wide flat tables, and cryptic names Unstructured documents, emails, logs Valuable Hidden connections and common business definitions
  6. 6. Graph Data Models & Semantics Simplifies access to complex data to address unanticipated questions Quickly profiles, connects and harmonizes data from multiple sources, including unstructured Presents tailored views and experiences to different personas with conceptual models Flexibly accommodates new data sources and use cases on the fly, with minimal impact Scales horizontally to accommodate enterprise data fabric scale
  7. 7. What it is ● An Enterprise Data Fabric Platform ○ Metadata Hub ○ Data Catalog ○ GraphMarts ○ Data Layers (where graph data blending happens) ○ REST Query Service Endpoints ○ Hi-Res Graph Aware Dash-Boarding Tool What it does ● Accelerated Data Integration as Services ○ Creates and stores a vast metadata description of the enterprises data landscape ○ Creates and stores metadata describing the transformations required to turn all raw data sources into a well described Enterprise Knowledge Graph ○ Automates the on-demand creation of the portions of the Knowledge Graph and brokers query access to it
  8. 8. What it is ● The first Graph Data Warehouse ○ GOLAP (Graph Online Analytics Processing) ○ In-Memory Massively Parallel Processing (MPP) ○ Linear Scale (Largest cluster 200x64 CPU servers) ○ Like Snowflake or AWS Redshift, but for Graph ○ Enterprise Scale Knowledge Graphs What it does ● Accelerated Data Integration & Analytics ○ ELT & Data Virtualization (VKG) ○ Knowledge Graph ingested from data sources using > 200 data source connectors ○ Reporting and BI analytics & aggregates ○ Graph Algorithms e.g. Page Rank, Shortest Path ○ Data Science libraries & Feature Engineering Transformations e.g Matrices, PCA, SVD ○ Labeled property graph (LPG) ○ Inferencing, windowed aggregates & views Standards • Supports Open Standards Supports RDF* and SPARQL 1.1 standards Where you can deploy • Fully Automated Deployment on premises or on cloud with Kubernetes Operator
  9. 9. Automation with Kubernetes
  10. 10. Kubernetes / K8s
  11. 11. Kubernetes API & AnzoGraph Operator
  12. 12. Kubernetes API & AnzoGraph Operator Kubernetes Cluster
  13. 13. Kubernetes API & AnzoGraph Operator Kubernetes Container Kubernetes Cluster
  14. 14. Kubernetes API & AnzoGraph Operator Kubernetes Container Cluster
  15. 15. Kubernetes API & AnzoGraph Operator
  16. 16. Kubernetes API & AnzoGraph Operator
  17. 17. Kubernetes API & AnzoGraph Operator
  18. 18. Kubernetes API & AnzoGraph Operator
  19. 19. Kubernetes API & AnzoGraph Operator
  20. 20. Kubernetes API & AnzoGraph Operator
  21. 21. Who supports Kubernetes?
  22. 22. The Kubernetes API provides a common automation abstraction across all cloud providers as well as on-premises implementations which allow us to deliver a hybrid multi-cloud deployment model for Anzo Enterprise Data Fabric with very low switching costs. Because all data transformation mappings, graph linking & blending instructions and all computing configurations are held as metadata in Anzo, customers can decide both when and where to deploy their data integration and analytics computing at the most granular level. This allows them to take advantage of the best available pricing and to more easily keep some workloads (and their data) behind their firewalls.
  23. 23. Thank You

×