As access to information, entertainment and communications edges closer to instantaneous, consumer expectations and demands for fast and seamless customer experiences grow as well. This is especially true for e-commerce businesses that cater to online customers who put a premium on convenience.
Many of the databases that support today’s e-commerce sites are built on MySQL. These sites typically function quite effectively until workload hits a critical mass, at which point disaster may strike. You can ‘imagine’ a rapidly growing retail site that has experienced 50 percent growth over a year period leading up to Cyber Monday. And at Magento Imagine 2015, we met literally dozens of e-commerce merchants that fit that description! And each time, their site’s database would usually have the capacity to handle a gradual increase in customers over the course of 12 months… But when they hit a sudden surge, for instance a flash sale, their site(s) slowed down to unusable levels, if not literally crashed.
“But I don’t do flash sales!” you may think. Well, if your provider develops an inventory problem, or one of your products suddenly blows-up on social media, you might just find yourself in a ‘flash sale’ situation… and crossing your fingers that your e-commerce site can handle the peak.
Or how about the infamous ‘Cyber Monday’? On the biggest online shopping day of the year a large percentage of your growing customer base will visit your site all at once, and all of a sudden the database is overwhelmed, leading to slower page loads and perhaps even site downtime.
Either of these outcomes can be catastrophic. Speed and performance are particularly important for e-commerce businesses; consider, for instance, that 64 percent of consumers who experience poor performance will abandon a site. Downtime is even more damaging than poor performance, as research shows that SMBs stand to lose $12,500 per hour that a site is down.
The following three case studies explore e-commerce sites which were faced with scalability challenges and how each was able to meet them:
- Nomorerack: The online flash-sale company increased revenue by an incredible 1,023 percent in 2012 and needed a way to maintain site availability and performance as growth continued. In 2013, Nomorerack turned to ClustrixDB and was able to achieve a staggering 600 percent explosion in online sales by scaling from 48 to 112 cores to handle a three-fold traffic increase. The e-commerce company broke sales records again in 2014 running ClustrixDB, once again seamlessly handling a three-fold traffic spike.
- Photobox: Europe’s leading online photo service was looking to improve availability, ease-of-management, capability and performance over their existing MySQL architecture, which had started to falter. Leveraging ClustrixDB’s scale-out architecture Photobox was able to significantly decrease downtime and performance problems and meet its entire range of potential challenges, including creating a completely fault-tolerant database.
- iOffer: The leading social commerce site where consumers bid for offerings was suffering from complicated application logic and rapidly increasing uneven loads across master and slave systems but wanted to avoid sharding—the process of splitting information across multiple databases—which can make managing data extremely complex. ClustrixDB helped iOffer avoid this process and assisted Nomorerack and Photobox in achieving their e-commerce site performance goals.