''the speed of data processing is just incredible; analytic query processing operations that used to take 30-45 minutes are now done within just 5 or 6 seconds. The EXASOL solution makes real-time analytics a reality.”
Sava Zibler, head of IT department, Mladinska knjiga Založba
Number of employees: 300
10 TB Exasol, Tableau, Python, R
Somewhat of a pioneer, Mladinska knjiga Založba, d. d. (MKZ) is the first company in Slovenia to deploy EXASOL, the fast in-memory analytic database. MKZ set out to deploy a data warehouse and introduce business analytics into its business in 2010; back then the company’s hardware platform was made up of IBM servers and data was processed by Netezza. Over recent years, the amount of data MKZ has been processing has increased dramatically, as has the number of business users who have chosen to look for analytical solutions in order to solve their business problems. However, MKZ’s previous analytical solution just could not cope with the growth in demand (some 600GB of data, 100 business users), so the company went out into the market to find a more innovative solution.
“The nights were too short and process operations were too onerous,” is how Igor Pušnar, head of analytics solutions development at MKZ, remembers the past. With its partner CRMT, the company examined various options and chose to test in-memory analytic solutions from five different technology providers. “Overnight, our data pool fills up with fresh data from all kinds of sources. Then, pre-set reports and data cubes are created. The volume of data – which is manageable – is not the main challenge. Instead, our previous solution had difficulties coping with the complexity of the data; we deal with a wide variety of data, sometimes featuring more than a hundred attributes, and a common denominator has to be found for various data types. The greatest challenge is executing complex and demanding ad hoc inquiries, something which cannot be standardized,” adds Pušnar.
MKZ studied the products available on the market, calling on an experienced database administrator for his input. He immediately recommended the EXASOL analytic database engine, noting that it would significantly reduce DBA overhead as the database is self-optimizing and automatically performs a large number of administration tasks itself.
As a result, MKZ introduced the EXASOL solution on a trial basis and was astounded from the very first time they used the database. Test data processing, which took an hour and fifteen minutes to complete in the incumbent analytic system, ran in EXASOL in just two seconds. MKZ also discovered another major benefit of the in-memory solution in the system’s economic use of resources. Thanks to table and data compression technologies, the solution requires only ten percent of the data volume to be kept in memory, which means MKZ only needs 60GB RAM capacity for processing 600GB of data.
“When we deployed the EXASOL database, our IT department had to do very little to make it perform. Our partner customized our ETL and syntax, which was a one-off task, and we could start to see distinct advantages with the system. And for business end-users, there was no impact, except for the fact that they suddenly had much faster reporting and data visualizations, with EXASOL delivering information to their MicroStrategy BI platform in a fraction of the time when compared to the previous solution. We estimate that full solution migration will require little more than fifty man-days in 2017,“ adds Sava Zibler, head of the IT department at MKZ.
At MKZ, business analytics are being deployed across an increasing number of departments and employees. Indeed, by the end of 2017, all co-workers who are responsible for taking important business decisions will be given full access to the new system.
“In financial terms, the investment in EXASOL may be comparable to the cost of maintaining the previous solution, but we get a much higher quality of service. What’s more, the fact that EXASOL is easier to use is a decisive factor. And the speed of data processing is just incredible; analytic query processing operations that used to take 30-45 minutes are now done within just 5 or 6 seconds. The EXASOL solution makes real-time analytics a reality,” says Zibler.
Given that the company stores and processes a large amount of personal customer data and needs to safeguard it from potential abuse, a cloud solution was ruled out as an option. As a result, MKZ performs all in-memory analytics operations on-premise using EXASOL on its servers in-house.