Komli Media, Asia Pacific’s leading media technology company, depends on reaching targeted audiences efficiently and at scale in order to please their customers. Over time, the company has collected more than 100 TB of data with information such as consumer behavior, clicks, impressions, and ads created. Users access this data in order to optimize campaigns to the consumer, adapt bidding algorithms in real-time, and find new advertising opportunities.
When Komli first implemented Hadoop, it had difficulty keeping up with demand because there wasn’t enough flexibility to adapt to the growing number of queries. As a result, the company had to turn away some requests. In addition, the amount of data Komli was collecting, 700 GB of raw data each day, had an incredibly slow turnaround time, and the cost for monthly cluster processing averaged at $15,000 a month.
Due to these challenges, Komli turned to Qubole for its true auto-scaling. With Qubole Data Service (QDS), a cluster will automatically expand when a load is high and cut back on nodes when the load decreases. This not only offers significant cost savings to the customer but also makes it easy to adapt to varying workloads without sacrificing performance. Other providers only offer manual auto-scaling, which requires the user to pre-define the auto-scaling capacity. QDS also offered higher performance than Amazon Web Services because of QDS’ I/O optimization for S3.
As a result of turning to QDS, Komli Media saw a significant decrease in processing time at a much lower price. Overall the company saw:
- 30% decrease in job processing times (waiting 1-2 hours for data instead of 10-15).
- 50% decrease in monthly processing costs.
- Ability to scale to meet all user requests.
“Our business has a lot more flexibility now thanks to QDS,” says Shailesh Garg, engineering manager at Komli. “Nowadays users get their data processed in one or two hours max instead of the 10 to 15 hours it took before QDS. I’m no longer afraid when my business users ask for more data. In fact, I’m happy that they are able to depend on the availability of Big Data to make better decisions and I encourage them to experiment with even more business questions.”