The new generation of databases differs from traditional databases in that they can quickly process the main types of data with massive simple data modeling capabilities. For TB-level data (or even PB-level data ), it can be easily handled; for our end users, real-time SQL statement operations are also required. In this case, the new-generation database standards must include: MPP-
The new generation of databases differs from traditional databases in that they can quickly process the main types of data with massive simple data modeling capabilities. For TB-level data (or even PB-level data ), it can be easily handled; for our end users, real-time SQL statement operations are also required. In this case, the new-generation database standards must include: MPP-
The new generation of databases differs from traditional databases in that they can quickly process the main types of data with massive simple data modeling capabilities. For TB-level data (or even PB-level data ), it can be easily handled; for our end users, real-time SQL statement operations are also required. In this way, the new generation of database standards must include:
- MPP-Large-scale parallel processing capability. The new generation of databases must provide large-scale parallel processing capabilities to achieve simultaneous import, processing, or query operations on different machines. The performance must be far better than that of traditional databases that only run on a single machine and have a data import bottleneck.
- Non-shared architecture: the non-shared (memory or disk) architecture means that the new generation of database systems will be impeccable! Each node works independently, so that other machines can continue to run even if one of them fails. This is very important in the MPP environment, because hundreds of machines may process data in parallel at the same time, it is difficult to avoid mechanical faults.
- Columnar structure: instead of queue-type data storage or processing that is common in relational databases. Under the columnar structure, only data in the column that plays a decisive role in the results will be processed. The results will be instantly presented by comparing and processing all the data. This also means that data is not organized into a neat list like in traditional databases.
- Excellent data compression capabilities: Excellent data compression capabilities allow the next generation of databases to consume and store more data, and save hardware resources more than traditional databases. A data warehouse with a 10:1 compression capability. For example, 10 TB data can be compressed to 1 TB. Data Compression is related to file encoding, which plays a decisive role in data compression.
- General hardware: Like Hadoop clusters, most of the new-generation databases can be used on general hardware facilities, from Dell, IBM, or others, in this way, users can make appropriate decisions after comparing a series of vendors.
Although the new database is just a blueprint, it is not difficult to look at the database development history to find that it may not be long before it will surpass the above standards.
Original article address: five standards for the new generation of databases. Thank you for sharing them.
