Camunda 8 Community version. What does it entail?

Camunda 8 Topics Discussion & Questions
9

@Niall Here is something I had listed for my own knowledge

Something that can be validated by your team and updated accordingly for more accuracy
Camunda 7 Camunda 8
1 Core Process Engine
  • Standalone Remote or embedded process Engine
  • Run complete camunda process engine in individual servers >2 behind load balancer with shared database
  • Zeebe has built in clustering with default 3 nodes in cluster
2 Webapp or User Interface
  • cockpit for process and task level monitoring and operational dashboards
  • Task list to start process, Manage tasks, create task filters etc
  • Process level views to manage process definition and process instances 
  • Cockpit provides basic process and task level dashboard 
  • Webapp(Optimize, tasklist and operate) are paid and not free in community edition 
  • However these are free in DEV and nonprod environments for free
  • Incident management through operate
  • Operational dashboards in Optimize
3 Modelling(BPMN, DMN and Forms) - Design time
  • Supports only Desktop modeler
  • No version management or collaboration 
  • Extensive support of BPMN 2.0 notations
  • Supports Web modeler(Paid) and as well as Desktop modeler(Community)
  • Version management and collaboration on BPMN/DMN/FORMS for better design experience
  • Does not support few BPMN 2.0 concepts like script task, few event gates etc
4 Modelling(BPMN, DMN and Forms) - Deploy time
  • Can be done via REST API or from Modeler in case of Remote Engine
  • In case of Springboot embeded all models needs to be placed in resources folder
  • Deploy models using grpc or desktop modeler and run using external task
  • Can be deployed via new web modeler as well
5 Tasklist - Task Management and monitoring - Runtime

6 Process Management and Monitoring  - Runtime

7 Process engine Deployments
  • Allows us to define the configuration and customize the process engine configuration 
  • Single package deployment either with embeded springboot or tomcat distribution or as base docker image
  • SAAS Offering : Leverages Camunda cloud Engine is deployed to kubernetes using helm charts provided by Camunda team
  • Self Managed Deployment custom helm charts and kubernetes configs provided by camunda team post purchasing license. Limited option to customize or overwrite the default settings
  • Everything is deployable as docker image and no integration to J2EE servers like tomcat
8 API integration 
  • Provides REST API endpoints and open API spec documentation around all APIs so all types of client(Node JS, Python, Java etc) can invoke these endpoints 
  • Does not support REST but does support gRPC and hence client needs to generate the code to invoke Engine APIs
9 Engine Plugins
  • Allows to customize process engine behavior before or after engine initialization 
  • Process engine can be customized in zeebe using exporter functionality but it is very limited
10 Connectors
  • Supports HTTP and SOAP connectors along with you can also extend connectors
  • Supports REST and SEND Grid Connectors(Email)
  • Cloud Connectors 
    • Example : Kafka, SQS, Apache Camel, AWS Lambda, Event Bridge
  • Service Connectors
    • Example : RPA, AI, IOT 
  • Content Connectors
    • Example : Open Test
  • Data Connectors
    • Example : BI systems, Datalakes or data ware house 
11 Operational dashboard and Search- Runtime View

12 Process and task events - Runtime

13 Authentication - Design/Runtime

14 Authorization - Design/Runtime

15 Multi tenancy

16 Reporting and Analytics - Runtime

17 Data Types
  • Allows both primitive and JSON data types to be stored as process variables along with Complex Java objects
  • SPIN library is used to parse XML/JSON data types within model 
  • DMN Supports both JUEL(Java Unified Expression Language) and FEEL(Friendly enough Expression language) 
  • Supports only JSON data type
  • Does not support SPIN library need to use our own transformation code for JSON/XML parsing
  • DMN Supports only FEEL 
18 State Management State data is stored in shared relational Database tables as records and all process engines in cluster are pointing to same database State data is stored as event stream locally within disk and is replicated across brokers
19 Scaling APP side can be scaled horizontally where as database side had limitation since DB was shared in clustered set up Since database is not bottle neck anymore Process engine can scale horizontally
20 DR set up Database replication is required across regions Brokers can replicate data much faster
4 Likes