One of the benefits of coming to a greenfield job – like when I joined Mind Candy two years ago – was that you can jump several technological steps ahead as you don’t have any legacy to deal with. Essentially we could build from scratch based on lessons learned from traditional data architecture. One of the main ones was to establish a real-time path right away to avoid having to shoehorn it in afterwards. Another was to avoid physical hardware. And the most important one was to hold off on Hadoop as long as possible.
The last one might seem surprising, isn’t Hadoop the centre-piece of a data architecture? Unfortunately it creates a lots of admin overhead and it might be a full person’s (or more) workload to maintain. Not ideal in a small company where people resources are limited. AWS S3 can fulfil most of the storage function but requires no maintenance and is largely fast enough. Also while HDFS is important and will probably come back for us soon, MR1 or YARN is just not – there are better and more advanced execution systems that can use HDFS and we used one of those: Mesos.
Mesos is a universal execution engine for job and resource distribution. Unlike YARN it can not only run Spark but also Cassandra, Kafka, Docker containers and recently also HDFS. That works because Mesos just offers resources and let the framework handle the starting and management of the jobs. This finally breaks the link between framework and execution engine: in Mesos you can run not only different frameworks but different versions of the same framework. No more waiting for your infrastructure to upgrade to the latest Hadoop or Spark version, you can run it right now even when all your other jobs run on older versions. Combine that with a robust architecture and simple upgrading and Mesos can easily be seen as the successor to YARN (for more details on why Mesos beats Yarn, see Dean Wampler’s talk from Strata).
For the real-time path the obvious processing solution is Spark Streaming (so we have a simpler code base) running on Mesos with Kafka to feed data in and with Cassandra to store the results. You now have a so-called SMACK stack (Spark Mesos Apache Cassandra Kafka) for data processing which the Mesos folks call Mesosphere Infinity for some reason (aka marketing).
The last bit of a data architecture is the SQL engine. Traditionally this was Hive but we all know Hive is slow. While there are several open-source solutions out there that improve on good old Hive (Impala, Spark SQL) in the end we decided on AWS Redshift. It’s a column-oriented SQL-based data warehouse with PostgreSQL interface which fulfils most of the data analysis and data science needs while being reasonably fast and relatively easy to maintain with few people.
The resulting architecture looks like the above picture. We have an event receiver and a enricher/validate/cleaner, which were written in-house in Scala/Akka and are relatively simple programs using AWS SQS as a transport channel. The data is then send to Kafka and S3. Spark uses data straight from S3 to aggregate and put the processed data back into either Redshift or S3. On the real-time side of things we have Kafka going into Spark Streaming with an output into Cassandra.
What can be improved here? HDFS is still better than S3 for certain large scale jobs and we want to bring it back running on Mesos. Redshift could be replaced with Spark SQL hopefully soon. All in all the switch from tightly coupled Hadoop to an open architecture based on Mesos allowed us to have an unprecedented freedom as to which kind of data jobs we want to run and which frameworks to use, allowing a small team to do data processing in ways previous only possible on a large budget.
I’m giving a talk about Spark Streaming and probabilistic data structures this Monday at the London Hadoop Meetup. Sign up with the link below!
Looking again at the data science diagram – or the unicorn diagram for that matter – makes me realize they are not really addressing how a typical data science role fits into an organization. To do that we have to contrast it with two other roles: data engineer and business analyst.
What makes a data scientist different from a data engineer? Most data engineers can write machine learning services perfectly well or do complicated data transformation in code. It’s not the skill that makes them different, it’s the focus: data scientists focus on the statistical model or the data mining task at hand, data engineers focus on coding, cleaning up data and implementing the models fine-tuned by the data scientists.
What is the difference between a data scientist and a business/insight/data analyst? Data scientists can code and understand the tools! Why is that important? With the emergence of the new tool sets around data, SQL and point & click skills can only get you so far. If you can do the same in Spark or Cascading your data deep dive will be faster and more accurate than it will ever be in Hive. Understanding your way around R libraries gives you statistical abilities most analysts only dream of. On the other hand, business analysts know their subject area very well and will easily come up with many different subject angles to approach the data.
The focus of a data scientist, what I am looking for when I hire one, should be statistical knowledge and using coding skills for applied mathematics. Yes, there can be the occasional unicorn in a very senior data scientist, but I know few junior or mid-level data scientist who can surpass a data engineer in coding skills. Very few know as much about the business as a proper business analyst.
Which means you end up with something like this:
Data scientists use their advanced statistical skills to help improve the models the data engineers implement and to put proper statistical rigour on the data discovery and analysis the customer is asking for. Essentially the business analyst is just one of many customers – in mobile gaming most of the questions come from game designers and product designers – people with a subject matter expertise very few data scientists can ever reach.
But they don’t have to. Occupying the space between engineering and subject matter experts, data scientists can help both by using skills no one else has without having to be the unicorn.
At a recent CloudCamp I had a discussion about data retention in the cloud, the argument was that the size of “big data” would be significantly reduced if you delete the unimportant/unnecessary/trivial data.
Problem 1: The Filtering Job
If you want to avoid collecting any unimportant data, it has to be filtered when coming in. If that would be an easy job, some companies would not use big data solutions – it would be less cost and resource intensive to just put it into a SQLDB. One of the reasons that it is necessary to work with cloud and big data solutions is that it is easier/less resource consuming to process the data later when you want to analyse it then when you receive it.
Problem 2: The Purging Job
If you can’t reasonably filter that data, how about purging it? All boils down to storage cost vs purging cost. If it is simple and effective to purge, you could have done it via filtering. It it’s not, you either have to spend precious resources for purging calculation or hire people to evaluate and purge data. Either way, it’s most likely more expensive than some more hard drives.
Problem 3: The Future
What is unimportant? What do you not need? If you think just about now, it might be an easy questions. But requirements might change, Data needs to be reprocessed in a different light. Your company might do something completely different with the data in a year (happens more often than you think). So why delete something you might need in the future?