Apache Spark is a multi-language engine for executing data engineering, data science, and machine learning on single-node machines or clusters.
N/A
Scala
Score 6.0 out of 10
N/A
Scala in Malvern, PA offers their digital signage software which provides Designer for content design, Content Manager for content organization and control, and Player for content viewing. Notably the software supports a wide array of digital signage including touchscreen kiosks and service for direct customer engagement and interaction.
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.
If you are in the data science world, Scala is the best language to work with Spark, the defacto data science data store. I think that is really the main likely reason I would ever recommend Scala. Another reason is if you already have a team of programmers familiar with functional programming, e.g. they all have years of Haskell experience. In that case, I definitely think Scala is a superior and faster-growing language than Haskell and that picking up Scala after Haskell should be quick.
The built-in compiler, scalac, is sssssssssslllllooooowwwwww. I mean like, if you thought the Java compiler was slow, try Scala! The default compiler on my 12k line codebase takes 4 minutes to compile from scratch on my i7 quad-core machine. This can be mitigated through the paid solution of Hydra which compiles your code in parallel. Unfortunately, it's quite expensive and your legal department or finance department may not approve of it. But if they do, for me, it reduced my compile time down to 80 seconds, much more manageable.
Scala is not going anywhere and support for it is slowly dying. This is the main reason I would not choose Scala for my next company or project. Important Scala libraries such as secure social (which is used for OAuth, a major requirement of every web app) are hardly maintained. Another library that suffers from lack of updates is Slick, the database mapper. There aren't enough engineers working on it to even provide support for the new features that came out in Postgres 9.0 (e.g. JSONb). There is simply not enough of a community to drive Scala forward and keep 3rd party libraries up to date as Java world does it.
As a corollary of a stagnant community, hiring Scala developers is hard as well. Of the 30 backend engineers we've hired, only 3 came in already knowing Scala. And as I will mention below, this is a BIG problem because learning Scala is really tough.
The learning curve for Scala is very, very steep. Anecdotally, I came into my current company with strong Java experience. Java is the closest language to Scala but it took me 6 months before I stopped needing to pair program on easy tickets. It doesn't help that Scala has some weird syntax like Map[A, +B] and that it forces you to do functional programming.
If the team looking to use Apache Spark is not used to debug and tweak settings for jobs to ensure maximum optimizations, it can be frustrating. However, the documentation and the support of the community on the internet can help resolve most issues. Moreover, it is highly configurable and it integrates with different tools (eg: it can be used by dbt core), which increase the scenarios where it can be used
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.
The customer service team is very responsive and usually returns calls or emails within a couple of hours of placing a request or inquiry. Just about every rep I've spoken to has been very thorough and helpful, walking me through each problem and explaining the solutions in a way that's easy to understand.
Spark in comparison to similar technologies ends up being a one stop shop. You can achieve so much with this one framework instead of having to stitch and weave multiple technologies from the Hadoop stack, all while getting incredibility performance, minimal boilerplate, and getting the ability to write your application in the language of your choosing.
Negative: slow engineer onboarding. As I mentioned before, it took me 6 months to get up-to-speed on Scala and didn't need to bother more senior Scala engineers anymore for help with every ticket. That's hundreds of hours I wasted of myself and other engineer's time.
Positive: thread safety, no concurrency bug. The ROI on this one is really hard to calculate, but I do believe Scala has saved me hundreds of hours over the past few years by allowing me to never have to worry about deadlocks or race conditions. Scala is simply so safe we've never had race conditions within the JVM before.
Negative: third-party libraries aren't maintained so we have to fork and update them ourselves. As I mentioned before, we use Securesocial but it stopped receiving updates and there is simply no alternative to it. So, we forked it and put an engineer on it for a month to get it back up-to-date. What a waste of his time!
