Software-Defined Storage
What Is Erasure Coding?
Erasure coding (EC) is a fault-tolerant coding technology. It was first used in the telecommunication industry to solve the data loss problem during transmission. This mechanism segments the signals to be transmitted, add signal parity, and then associate the segments. Even if some signals are lost during transmission, the receive end can still calculate the complete information through an algorithm. In data storage, the erasure coding mechanism divides data into segments, extend and encode redundant data blocks, and store them in different locations such as drives, storage nodes, or other physical locations.
The erasure code is a structure of k data blocks and m parity blocks. The values of k and m can be set according to a specific rule. The formula is n = k + m. The variable k indicates the number of original data blocks. The variable m indicates the number of parity blocks. The variable n indicates the total number of blocks after the erasure coding mechanism is implemented. When less than or equal to m storage blocks (data blocks or parity blocks) are damaged, all data blocks can be obtained by calculating data in the remaining storage blocks, preventing data loss.
For more information about erasure coding, see Kunpeng BoostKit for SDS Technical White Paper.
What Is the Relationship Between glz and the Compression Algorithm?
The Kunpeng BoostKit for SDS compression algorithm is a Huawei proprietary lossless compression algorithm Compared with open-source compression algorithms, the Huawei-proprietary compression algorithm offers a higher compression rate and performance. It provides 25% higher compression rate and 10% higher bandwidth than mainstream open-source compression algorithms.
The name of the compression algorithm software package contains glz, for example, BoostKit-glz_1.0.2.zip. Therefore, the compression algorithm is also referred to as glz. The compression algorithm was previously named KPS Zip.
For more information about the compression algorithm, see Compression Algorithm Feature Guide.
What Are the Requirements of the KSAL on openEuler?
The Kunpeng Storage Acceleration Library (KSAL) supports openEuler 20.03 and openEuler 22.03.
For Ceph 14.2.8, the smart write cache feature supports openEuler 20.03 and openEuler 22.03.