Développez l'apprentissage automatique prédictif avec Flink | Atelier du 18 déc. | S'inscrire
"What is our role with Confluent Cloud?” is a valid question frequently asked by service and delivery partners who have traditionally made money from services related to the installation and upkeep of on-prem applications. To answer this, there are two key prerequisites that need to be understood:
What offerings does Confluent Cloud provide?
As a service partner, what additional activities need to be performed to support Confluent Cloud customers?
The aim of this blog is to provide guidelines / a checklist for service partners to fully utilize Confluent Cloud services and eliminate any gray areas related to scope of work that needs to be performed by service partners in a SaaS world.
Confluent Cloud (CC) is a fully managed cloud native Kafka service for connecting and processing all of your data, everywhere it's needed. We provide the following functions as a service:
Connectors
Kafka
ksqlDB
Schema Registry
Other enterprise-ready features include, but are not limited to:
Monitoring (Control Center)
RBAC
Audit Logs
Data Flow for Data Lineage
Enterprise Support
All services can be self-provisioned based on customer needs and are available on all major cloud providers such as AWS/Azure/GCP, with a pricing model based on consumption/Usage Based Billing (UBB).
If you’re asking, “What is a fully managed service?” We take care of:
Installation and management
At-rest & in-transit data encryption
Uptime SLA’s
Throughput guarantees
Upgrades
And much more!
Let me draw an analogy for better understanding: the engine of a Formula 1 car is the most critical element and is typically built by a specific engine manufacturer. In turn, it’s sold to and used by multiple racing teams.
Consider a scenario where engines built by Mercedes are put into Mercedes, McLaren, Aston Martin, and Williams cars.
As an engine expert and manufacturer, Mercedes ensures the performance of the engine alone and takes responsibility for the engine, while the racing car companies take ownership of the rest of the car like design, weight, wheels, aerodynamics, etc.
This is exactly what Confluent Cloud offers: our technology is the engine that racing car companies (service partners) use to build multiple cars (applications), all while using the same engine.
Confluent Cloud provides the engine while the race cars are built by Service Partners!
The benefit to the end customer is the ability to build multiple data streaming applications quickly, all while benefiting from running on Confluent Cloud.
To my curious reader's key question, “What activities need to be performed by me as a service provider on Confluent Cloud?”
I’d like to break up activities into three different personas:
Developers
Operators
Architects
This involves the development of applications on Confluent Cloud as well as the configuration of Confluent to support these applications or use cases in an optimal way. This typically entails:
Connectors
Choosing the right connector
Optimizing the number of tasks
Configuring Single message transforms (SMTs)
Meta-data information
ksqlDB
Scoping the number of CSUs required
Developing ksql pipelines
Kafka
Estimating the number of topics
Topics creation with partitioning strategy
Data retention/data replication
Data format
Schema registry
Creating schemas with compatibility modes
All of this must be designed and built by our service partners.
As you can see, the developer’s activities remain the same be it a self managed Confluent or Confluent Cloud.
As you might have expected, there are some tasks that are generally associated with on-prem operations which are omitted when using SaaS software like Confluent Cloud. These are duties such as software installation, configuration, support, etc. These are provided as out of the box on Confluent Cloud. What remains is still important:
Monitoring - Control Center for Confluent Cloud provides a vast collection of metrics related to the health of the environment. For use cases where additional metrics need to be monitored, Confluent Cloud exposes the Metrics API which can be integrated into a 3rd-party monitoring tool of choice.
Application Deployment - Deployment of applications built into higher environments on Confluent cloud, application change management life-cycle, etc.
Application Support - Applications built by service partners on top of Confluent Cloud have to be supported by service partners for their lifetimes.
The majority of Architect activities remain the same except for some activities like cluster sizing, security, etc., which are not provided as out of the box by Confluent Cloud. The remaining tasks are vitally important:
Architecting the overall solution by defining the type of clusters required, replication and HA strategies, capacity, DR considerations, Schema Registry, etc.
Security - Setting up of security i.e. authentication and authorization activities such as Single Sign On, adding users, defining roles, ACLs, API key generation and management, etc.
Performance Benchmarking - Confluent Cloud provides max read and write throughput at the cluster level, each cluster will have more than one topic which actually holds all events. Topic design becomes critical to ensure Topics can scale up to individual throughput requirements.
With a list of things to do discussed above, let me come to the easy section and tell you what doesn’t need to be done on Confluent Cloud:
Identifying the infrastructure i.e. defining number of brokers, ZK, memory, CPU, etc.
Installation and upgrades of software services such as Kafka, connectors, ksqlDB, etc.
Infrastructure support of software services
To summarize, there is minimal impact on the scope of activities for the Developer and Architect, while some of the Operator activities have been automated and provided as out of the box by Confluent Cloud.
Conclusion Confluent Cloud provides a platform whose stability is underpinned by years of operational expertise by the founders and majority contributors of Apache Kafka, while also adding:
ELASTIC - massive scale when required
GLOBAL - built for hybrid and multi cloud environments in all major clouds
INFINITE - infinite data storage
COMPLETE - Not only Kafka, connectors, ksqlDB, Schema Registry, and much more!
HIGHLY AVAILABLE - 99.9% Kafka SLA
SECURE - Enterprise-ready encryption and security features out of the box
So while we provide a best in class technology, service partners have a major role to play in bringing in:
Deep understanding of customer landscape
Domain knowledge
Application development expertise
Integration with other technologies
To help customers realize the full business value of Confluent Cloud in a quick time.
This blog explores how cloud service providers (CSPs) and managed service providers (MSPs) increasingly recognize the advantages of leveraging Confluent to deliver fully managed Kafka services to their clients. Confluent enables these service providers to deliver higher value offerings to wider...
With Confluent sitting at the core of their data infrastructure, Atomic Tessellator provides a powerful platform for molecular research backed by computational methods, focusing on catalyst discovery. Read on to learn how data streaming plays a central role in their technology.