Big Data Management Solutions For Large, In-session Gaming Data

As a wholly owned subsidiary IGT Canada Solutions ULC, based in Moncton NB, employs over 450 staff anddevelops gaming products for global markets. IGT is faced with huge gaming data (one month's worth of datacould go up to 1 TB). In the era of Big Data, data from a global customer population (wager, cash in amount,cash out amount, and win amount) is seen as an asset allowing better understanding of the market needs andopportunities. Global gaming data that is time consuming to transmit and difficult to handle requires a datamanagement solution that will scale, while at the same time accommodating the requirement for more insightinto gaming sessions and allow IGT to provide answers to many questions that will give it a betterunderstanding of its business in different global markets. Given that the data sizes exceed the capacities of SQLdatabases, one turns to no-SQL solutions. We envision that the tools tested will be a non-SQL database calledMongDB or Cassandra, and a non-SQL index called Solr.We want to investigate which of these approaches may offer some advantage to IGT data search.We need to know the scaling of data management to datasets growing incrementally at the level of 1TB amonth, s well as the speed to these data storage techniques for different on-session analytics queries.This research is focused on measuring data retrieval time for some queries defined by IGT. Considering thelimitations of the query languages available for non-SQL products, we focus the measure of how much time isconsumed to retrieve the data for simple queries. If IGT can query its data efficiently it will allow the companyto explore new evidence-based market opportunities.