Need advice about which tool to choose?Ask the StackShare community!
Apache Spark vs Singer: What are the differences?
Introduction:
Apache Spark and Singer are two popular tools in the data processing domain. While both serve similar purposes, they have key differences that set them apart. Below are the main distinctions between Apache Spark and Singer.
Processing Framework: Apache Spark is a distributed computing system that can process large datasets across multiple nodes in a cluster, making it ideal for big data tasks. On the other hand, Singer is an open-source ETL (Extract, Transform, Load) tool that focuses on extracting data from various sources, transforming it, and loading it into desired destinations. While Spark is more versatile in terms of processing capabilities, Singer is specifically designed for data integration tasks.
Programming Language: Apache Spark primarily uses Scala, although it also supports programming in Java, Python, and R. Spark provides APIs for these languages, making it accessible to a wide range of developers. Singer, on the other hand, relies on a configuration file written in JSON to define data extraction and transformation tasks. This makes Singer easier to use for individuals with less programming experience.
Real-time Processing: Apache Spark is known for its ability to handle real-time data processing, thanks to its streaming capabilities using technologies like Spark Streaming and Structured Streaming. On the flip side, Singer is more suited for batch processing, where data is processed in large batches rather than in real-time. If real-time processing is a critical requirement for a project, Apache Spark would be the preferred choice over Singer.
Supported Data Sources: Apache Spark supports a wide range of data sources and formats, including traditional databases, data lakes, streaming data sources, and more. It can seamlessly integrate with various data storage solutions, making it a versatile choice for diverse data processing needs. In contrast, Singer specializes in extracting data from API endpoints, databases, and other sources using a pre-built set of taps, limiting its flexibility compared to Spark's broader data source support.
Scalability: Apache Spark is built for scalability, allowing users to scale their processing tasks horizontally by adding more nodes to the cluster. This distributed computing approach enables Spark to handle massive datasets efficiently. Singer, while capable of running on scalable infrastructure, may not offer the same level of scalability as Spark due to its focus on ETL orchestration rather than distributed data processing.
Community and Ecosystem: Apache Spark boasts a robust community and a vast ecosystem of tools and libraries that extend its capabilities for various use cases. This includes machine learning libraries like MLLib and graph processing capabilities with GraphX. Singer, while having an active community, may not have as extensive an ecosystem as Spark, limiting its adaptability for complex data processing tasks that require specialized tooling.
In Summary, Apache Spark excels in distributed computing, real-time processing, and scalability, making it ideal for big data and complex processing tasks, whereas Singer specializes in ETL tasks, particularly for data extraction and transformation from various sources in a batch processing environment.
We have a Kafka topic having events of type A and type B. We need to perform an inner join on both type of events using some common field (primary-key). The joined events to be inserted in Elasticsearch.
In usual cases, type A and type B events (with same key) observed to be close upto 15 minutes. But in some cases they may be far from each other, lets say 6 hours. Sometimes event of either of the types never come.
In all cases, we should be able to find joined events instantly after they are joined and not-joined events within 15 minutes.
The first solution that came to me is to use upsert to update ElasticSearch:
- Use the primary-key as ES document id
- Upsert the records to ES as soon as you receive them. As you are using upsert, the 2nd record of the same primary-key will not overwrite the 1st one, but will be merged with it.
Cons: The load on ES will be higher, due to upsert.
To use Flink:
- Create a KeyedDataStream by the primary-key
- In the ProcessFunction, save the first record in a State. At the same time, create a Timer for 15 minutes in the future
- When the 2nd record comes, read the 1st record from the State, merge those two, and send out the result, and clear the State and the Timer if it has not fired
- When the Timer fires, read the 1st record from the State and send out as the output record.
- Have a 2nd Timer of 6 hours (or more) if you are not using Windowing to clean up the State
Pro: if you have already having Flink ingesting this stream. Otherwise, I would just go with the 1st solution.
Please refer "Structured Streaming" feature of Spark. Refer "Stream - Stream Join" at https://spark.apache.org/docs/latest/structured-streaming-programming-guide.html#stream-stream-joins . In short you need to specify "Define watermark delays on both inputs" and "Define a constraint on time across the two inputs"
Pros of Singer
- Multiple inputs "taps"1
- Open source1
Pros of Apache Spark
- Open-source61
- Fast and Flexible48
- One platform for every big data problem8
- Great for distributed SQL like applications8
- Easy to install and to use6
- Works well for most Datascience usecases3
- Interactive Query2
- Machine learning libratimery, Streaming in real2
- In memory Computation2
Sign up to add or upvote prosMake informed product decisions
Cons of Singer
Cons of Apache Spark
- Speed4