Apache Spark vs. Azure HDInsight vs. Hortonworks Data Platform

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

Well suited: A tiny-mid sized company with no immediate plans of growing the volume of their data processing, that can afford long response times from support. Also it helps if you are not prone to put your hands on Linux and Spark configuration. In fact, it can make things go really faster if you also work with the bundle-in Jupyter. And, if you need to perform some diagnostics and / or administrative tasks, that's full of tools to find an understand the Root Cause. Ideal for non experts. Less appropriate: Big Data company, intense on demand cluster creation, mission critical, costs reduction, latest versions of libraries required, sophisticate customizations required.

Incentivized

I find HDP easy to use and solves most of the problems for people looking to manage their big data. Evaluating the Hortonworks Data Platform is easy as it is free to download and install in your cluster. Single node cluster available as Sandbox is also easy for POCs.

Incentivized

• Rich APIs for data transformation making for very each to transform and prepare data in a distributed environment without worrying about memory issues
• Faster in execution times compare to Hadoop and PIG Latin
• Easy SQL interface to the same data set for people who are comfortable to explore data in a declarative manner
• Interoperability between SQL and Scala / Python style of munging data

Incentivized

• Data is presented without interfering others (IT or other dept).
• Data is managed properly and is available for retrievable any time.
• Legacy use of CD/DVD and Pendrive are not required.

Incentivized

• It does a good job of packaging a lot of big data components into bundles and lets you use the ones you are interested in or need. It supports an extensive list of components which lets us solve many problems.
• It provides the ability to manage installations and maintenance using Apache Ambari. It helps us in using management packs to install/upgrade components easily. It also helps us add, remove components, add, remove hosts, perform upgrades in a convenient manner. It also provides alerts and notifications and monitors the environment.
• What they excel in is packaging open source components that are relevant and are useful to solve and complement each other as well as contribute to enhancing those components. They do a great job in the community to keep on top of what would be useful to users, fixing bugs and working with other companies and individuals to make the platform better.

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

• The only problem I have come across is when loading large volumes of data I sometimes get an error message, I assume this means something is corrupt from within. I would love a way for this to be resolved without having to start over.

Incentivized

• Since it doesn't come with propriety tools for big data management, additional integration is need (for query handling, search, etc).
• It was very straightforward to store clinical data without relations, such as data from sensors of a medical device. But it has limitations when needed to combine the data with other clinical data in structured format (e.g. lab results, diagnosis).
• Overall look and feel of front-end management tools (e.g. monitoring) are not good. It is not bad but it doesn't look professional.

Incentivized

Capacity of computing data in cluster and fast speed.

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

Incentivized

Azure HDInsight is usable on the top of Azure Data Lake and gives us the benefit of analyzing large scale data workload in Hadoop. Usability and support from Microsoft are outstanding.

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.

Inexpert, isolated teams... not good for support an excessively complex platform. Lots of weeks or months for a complex problem troubleshoot. Many time lost stuck on MindTree, before the case was finally escalated with Microsoft!

Incentivized

Try not to change variable names.

Incentivized

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.

Incentivized

At this time I have not used any other similar products... I am open to it but Azure HDInsight and its components really work well for our organization.

Incentivized

We chose [Hortonworks Data Platform] because it's free and because [it] was an IBM partner, suggested as big data platform after biginsights platform.

You can install in more physical computer without high specs, then you can use it in order to learn how to deploy, configure a complete big data cluster.

We installed also in a cloud infrastructure of 5 virtual machine

Incentivized

• Business leaders are able to take data driven decisions
• Business users are able access to data in near real time now . Before using spark, they had to wait for at least 24 hours for data to be available
• Business is able come up with new product ideas

Incentivized

• ROI is of course there, as no legacy software for data presentation.
• No manual intervention for data retrieval.
• Data is available anywhere as requested.

Incentivized

• It is difficult to have a negative impact, because the required investment is not that high.
• The big open community behind Hortonworks and related Apache Project makes it easy to put 'the wheel to meet the road' quite quickly.
• We have seen management meetings where the attendants were impressed by the results achieved with the datalake built on HDP.

Incentivized