Overview
Usage Engine Private Edition now supports batch scaling, making it possible to increase or decrease processing capacity as needed without manual intervention. As a general concept, batch scaling is a way to speed up processes by splitting the workload between multiple “workers” or resources, enabling them to complete tasks in parallel rather than sequentially. Usage Engine’s solution consists of three new agents, A Scalable File Collection agent and Scalable InterWF Forwarder and Collector agents. A new profile has also been created - the Partition Profile. The feature uses the existing agents, Data Aggregator and Deduplication, which have been updated to include a Kafka storage profile. Kafka must be configured for all storage within your batch scaling solution. Add something here about recommended use cases as per the note above?
How it works
This example shows a batch processing setup where you collect files and perform duplication checks and aggregation. We want to make this solution scalable to improve the processing times of our data during periods of high usage. We need to set up two to three workflows in our batch scaling solution. In this example, we use three.
I think we should add something here to explain what a partition is… also may be helpful to link to this Doc..Automatic Scale Out and Rebalancing (4.3).
Partitions - From Chat GPT:
In software scaling, a partition is a way of breaking up large sets of data or tasks into smaller, more manageable parts. Each partition handles a subset of the total data or workload, which allows the system to process different parts at the same time, using multiple resources (like servers or processors). This makes the overall process faster and more efficient, especially as the data or workload grows.
For example, if you have a huge list of data to process, a partitioned system could split that list into sections and process each one in parallel, speeding up the work and allowing the system to handle more data without slowing down.
The File collection workflow(s) manage the Inter workflow (InterWF) partitions. They will use an ID Field (e.g. customer ID) to determine which partition a UDR belongs to.
The number of partitions created is determined by the Max Scale Factor parameter. This is configured in ….
Note!
The number of partitions will be the same across all storage buckets/caches/topics. The points of storage will occur, for example,
With the passing of UDRs between workflows.
When duplicate UDR keys are detected.
For aggregated sessions.
The Duplication Check workflow(s) will check for duplicates across all partitions. Checked UDRs are placed in an additional topic with the same partitions as the corresponding Collection workflow topic. (The Duplicate keys are saved in a separate topic with the same number of partitions having the same ID fields.)
The Aggregation workflow(s) will collect data from an inter-workflow topic and use a separate aggregation session storage topic.
Prerequisites for Kafka?
Are there any prerequisites required to be able to configure batch scaling using Kafka storage?
Subsections
This section contains the following subsections:
you use the new agent InterWF Collector, to pick up the files from the external system/ IF storage (InterWF partition). You also need to have Duplication checks after which you will use the InterWF Forwarder to take the non-duplicated files and feed them to the Aggregation partitions on the data (pretty common processes in any workflow group) You will use the current agents Deduplicate and Data Aggregator, however, they will have a new storage profile option for Kafka, which you need to configure. Finally you would use the other new agent
From Chat GPT re: Topics - For draft purposes only:
In a software context, especially in messaging and streaming platforms like Kafka, a topic isn’t a type of storage in the traditional sense, like a cache or database. Instead, it refers to a "channel" or "feed" where messages (like UDRs) are grouped and published for consumers to read from. While a topic involves data persistence (messages are stored temporarily or longer-term, depending on configuration), it's more about organizing and transmitting data rather than being a storage unit itself.
In comparison, a cache is a direct storage solution intended for fast access to data. Topics, on the other hand, are about managing and distributing data streams efficiently across systems.