Apache Flume vs. Apache Spark

Apache Flume is well suited when the use case is log data ingestion and aggregate only, for example for compliance of configuration management. It is not well suited where you need a general-purpose real-time data ingestion pipeline that can receive log data and other forms of data streams (eg IoT, messages).

Incentivized

Well suited: To most of the local run of datasets and non-prod systems - scalability is not a problem at all. Including data from multiple types of data sources is an added advantage. MLlib is a decently nice built-in library that can be used for most of the ML tasks. Less appropriate: We had to work on a RecSys where the music dataset that we used was around 300+Gb in size. We faced memory-based issues. Few times we also got memory errors. Also the MLlib library does not have support for advanced analytics and deep-learning frameworks support. Understanding the internals of the working of Apache Spark for beginners is highly not possible.

Incentivized

• Multiple sources of data (sources) and destinations (sinks) that allows you to move data form and to any relevant data storage
• It is very easy to setup and run
• Very open to personalization, you can create filters, enrichment, new sources and destinations

Incentivized

• Apache Spark makes processing very large data sets possible. It handles these data sets in a fairly quick manner.
• Apache Spark does a fairly good job implementing machine learning models for larger data sets.
• Apache Spark seems to be a rapidly advancing software, with the new features making the software ever more straight-forward to use.

Incentivized

• It is very specific for log data ingestion so it is pretty hard to use for anything else besides log data
• Data replication is not built in and needs to be added on top of Apache Flume (not a hard job to do though)

Incentivized

• Memory management. Very weak on that.
• PySpark not as robust as scala with spark.
• spark master HA is needed. Not as HA as it should be.
• Locality should not be a necessity, but does help improvement. But would prefer no locality

Incentivized

Capacity of computing data in cluster and fast speed.

The only thing I dislike about spark's usability is the learning curve, there are many actions and transformations, however, its wide-range of uses for ETL processing, facility to integrate and it's multi-language support make this library a powerhouse for your data science solutions. It has especially aided us with its lightning-fast processing times.

Incentivized

Apache Flume is open-source so support is limited. Never the less, it has great documentation and best practices documents from their end-users so it is not hard to use, setup and configure.

Incentivized

1. It integrates very well with scala or python. 2. It's very easy to understand SQL interoperability. 3. Apache is way faster than the other competitive technologies. 4. The support from the Apache community is very huge for Spark. 5. Execution times are faster as compared to others. 6. There are a large number of forums available for Apache Spark. 7. The code availability for Apache Spark is simpler and easy to gain access to. 8. Many organizations use Apache Spark, so many solutions are available for existing applications.

Apache Flume is a very good solution when your project is not very complex at transformation and enrichment, and good if you have an external management suite like Cloudera, Hortonworks, etc. But it is not a real EAI or ETL like AB Initio or Attunity so
you need to know exactly what you want. On the other hand being an opensource project give Apache a lot of room to personalize thanks to its plug-able architecture and has a very nice performance having a very low CPU and Memory footprint, a single server can do the job on many occasions, as opposed to the multi-server architecture of paid products.

Incentivized

All the above systems work quite well on big data transformations whereas Spark really shines with its bigger API support and its ability to read from and write to multiple data sources. Using Spark one can easily switch between declarative versus imperative versus functional type programming easily based on the situation. Also it doesn't need special data ingestion or indexing pre-processing like Presto. Combining it with Jupyter Notebooks (https://github.com/jupyter-incubator/sparkmagic), one can develop the Spark code in an interactive manner in Scala or Python

Incentivized

• Flume has simplified a lot many of our ingest procedures, easier to deploy and integrate than a classical EAI, reducing the time to market
• But opposed to EAIs if the project starts to grow in complexity Apache Flume project may not be as suitable

Incentivized

• Faster turn around on feature development, we have seen a noticeable improvement in our agile development since using Spark.
• Easy adoption, having multiple departments use the same underlying technology even if the use cases are very different allows for more commonality amongst applications which definitely makes the operations team happy.
• Performance, we have been able to make some applications run over 20x faster since switching to Spark. This has saved us time, headaches, and operating costs.

Incentivized