Improved method aimed at small files of HDFS
A technology of small files and files, applied in special data processing applications, program control design, instruments, etc., can solve problems such as low efficiency of HDFS
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0042] Such as figure 1 As shown, an HDFS improvement method for small files includes a cluster, which contains a Namenode and multiple Datanodes, which can be accessed by multiple clients, and part of the authority of the Namenode is delegated to the Datanode node, allowing the Datanode to cache Part of the metadata information of small files handles most of the small file read and write requests.
Embodiment 2
[0044] On the basis of Embodiment 1, in this embodiment, besides Namenode manages all the metadata of the file system, Datanode also saves part of the metadata, mainly the metadata information of small files, and the metadata information of large files is still stored on Namenode. Among them, Namenode is responsible for managing system-wide activities, and periodically communicates with each Datanode with heartbeat information, providing them with instruction operations and collecting their feedback status.
Embodiment 3
[0046] On the basis of Embodiment 1, the operation of the client in this embodiment assigns the metadata operations of some small files to the Datanode. If no corresponding result is found, it is searched on the Namenode. When writing files, the client follows the previous Read and write records, directly query whether the Datanode has a block file block that is not full and no other client is writing at this time, if there is, write the data directly to this data block, and update the corresponding metadata information, if not , then send a write data request to the Namenode, and then the Namenode allocates a new data block to complete the data writing, the client queries the data block is completed on the local machine; when reading the file, directly query the Datanode, if not found, then Find Namenodes.
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com