Lindorm: Reduce Storage Costs by 80% with Cloud-Native Time Series and Spatial-Temporal Database
Catch the replay of the Apsara Conference 2020 at this link!
By ApsaraDB
At the Apsara Conference 2020 held on September 18, Alibaba Cloud launched Lindorm, a Time Series and Spatial-Temporal Database. It provides wide table engines, time series engines, search engines, and file engines to redefine the storage mode for multi-type data in the Internet era. Lindorm can simultaneously store, query, and retrieve various types of data, such as key-value data, wide table data, time series data, files, and images. It solves the problems, such as complex architecture and difficult maintenance, high data storage cost, and difficulty in coping with flexible business scale, when different types of databases are deployed respectively. It helps reduce massive data storage costs by 80% and is the first choice for scenarios, such as Internet of Vehicles (IOV), advertising, social networking, and gaming.
Reportedly, the Alibaba Cloud Lindorm database has been tested by the Alibaba economy for ten years and supports core businesses, such as Taobao, Tmall, Alipay, Cainiao Network, and IoT, with a throughput of tens of millions per second, hundreds of petabytes of storage capacity, and single-digit millisecond response latency. Based on the cloud-native architecture and with the separation of storage and computing, it provides geo-disaster recovery and global data synchronization. It can be flexibly scaled according to the business scale to match the rapid growth of the business.
The multi-modal technology has become the trend of the times, and traditional databases are carrying out the multi-model extension. However, this extension is based on the original architecture, and different data models need appropriate technical architectures to support them. Therefore, although this extension supports the storage of new data models, functionality, performance, and scalability are limited and cannot truly match the business needs. The Lindorm database is based on the self-developed cloud-native multi-modal engines and cloud-native storage engines. It is specially designed for multi-model data and has ultimate performance and scalability, making the system architecture simple and efficient. It greatly reduces system maintenance costs and massive data storage costs.
Support Multi-Model Data and Simplify the System Architecture
In traditional solutions, multiple databases are required to process different types of data separately. This leads to a complex architecture that is difficult to maintain and scale out in the case of dynamic and elastic services. The data is redundant and wasted.
Lindorm has built-in wide table engines, time series engines, search engines, and file engines. Each engine shares storage and supports independent auto scaling to replace these databases, enabling low-cost data storage and complex processing.
Reduce Mass Data Storage Costs by 80%
Data is one of the most valuable assets of an enterprise. However, due to the high costs of data storage, enterprises have to discard large amounts of valuable data. Based on self-developed low-cost elastic storage media, intelligent cold and hot separation technology, and adaptive compression algorithms that are transformed by zero application, Lindorm reduces the storage costs of massive data by 80%.
It is reported that Lindorm can provide services for users through Alibaba Cloud and Apsara Stack. By Lindorm, many users are building orders, logs, risk control, recommendations, social networking, IOV, intelligent buildings, Node Package Management (NPM), Application Performance Management (APM), which covers dozens of industries, such as IoT, internet, finance, gaming, and industrial production.
We kept broad-minded, accumulate experiences, and finally succeeded. After ten years of refinement and successful business validation, Lindorm maintains Alibaba’s understanding of the value of enterprise data. It is committed to building low-cost storage and processing solutions for data of any size and type, to make enterprise data “affordable and visible”.
