- April 10, 2018
- Posted by: admin
- Category: Uncategorized
In today’s world, geographically distributed databases, while difficult to design, are challenging administrative processes. There are available solutions, but they are few and far between; and none of them ideal. Managing these decentralized nodes is a hard problem to solve, therefore compromises are being made. Those problems can be divided into:
Did you know that Amazon estimates that each 100ms of latency is added because of centralized databases thousands of miles away, costing them 1% in sales? Shardix is going to solve that issue by developing decentralized databases via blockchain technology by ultimately reducing the latency and keeping the shards of data close to end-users.
Centralized databases suffer from not only latency but also availability problems. Therefore a great amount of effort and money is being put into so-called Service Level Agreements (SLA’s) in order to try to increase the “nines” in uptime, ultimately trying to achieve “five nines” which means 99.999% uptime. The process of achieving those numbers requires replicating the same data across various geographically distinct nodes. That might result in increased latency, costs and overall inefficiency. Shardix team wants to solve that by deploying a fault-tolerant system with adjustable SLA guarantees (99.99%, 99.999% etc.)
All these problems mentioned previously come with tremendous costs, both because of having so many servers around the world and the expenditures of the facilities. One of the biggest cloud distributed storage providers — Amazon Oracle SSD, charges up to $.20 per gigabyte, per month for SSD storage. That can easily translate into $200 dollars per month or $2,400 dollars per year for a terabyte of solid-state storage. With the ever decreasing costs of SSD’s (which can be bought now under $500 per gigabyte) and very low maintenance cost of those drives, the overall price tag can be much lower versus the traditional datacenter hosted models. This is achieved by renting out unused capacity from existing servers.
All those factors lead to an opportunity for a geographically distributed, replicated database which combines fault tolerance and efficiency while keeping high overhead. For years, the developer community has failed to deliver, and the technology to implement a system like this has been lacking as well. Now a new architecture has emerged — blockchain, and with it — Shardix.