Adaptive concurrency for write persistence
A persistent, self-adaptive technology, applied in special data processing applications, input/output to record carrier, input/output process of data processing, etc., which can solve the problem that it is difficult to provide enhanced feature solid-state memory
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment Construction
[0022] In various embodiments described herein, the storage system has a write process selector that selects among multiple write processes to optimize writing for different write requests. Possibilities for the write process include writing data to NVRAM (Non-Volatile Random Access Memory) to be flushed to flash later in the background write process, writing to NVRAM outside of the inode lock, updating while the inode is locked inode and commit data, outside of inode lock, write directly to flash and bypass NVRAM, update inode and commit data while inode is locked. The choice of write process is based on whether the data is small or large relative to a threshold, or whether data arrives serially or in parallel, as different write processes have different efficiencies in these cases. The writing process uses locking so that the file's I-node remains consistent across concurrent writes between storage nodes and solid-state storage units. Data is persisted through write commits...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


