Lexst- Data-Cluster

Case Study of LEXST Database

ET-China's air ticket system

ET-China is a travelling agency chain in South China, every day ET-China downloads more than million rows of discount data from the their supplier, and retails air tickets to customers. Because they have so many outlets and distributors, which search the price info concurrrently during working hours, database is too "big"(although million row table is not big for LEXST), their SQL servers became very slow(10-15 seconds to return a query), and almost refused to work. It's of course unacceptable for sales department, they gave trememdous pressure upon Tech department find a faster database solution and stop telling customers "please wait for a while, just moment.".

Technical department knew that the only possible solution should be putting the whole database into cache memory instead of hard disk, it meant a much bigger budget to buy a minicom(small computer) with big RAM.

Et-china technicians got know LEXST system by accident, actually LEXST database that time had not yet published, the core part of the engine was just finished and tested. Becaues of the urgent need, engineers from both parties installed a few Lexst nodes and wrote a simply customized API connected to their terminal, the system becomes much quicker, and average responding time sharply reduced 30 times.

When LEXST team told Et-China later that better Version can be provided, and free of charge. To everyone's surprise, our "first customer" refused without any hesitation: "No, no upgrade, thanks god, it's working pretty well now."