Multi-core storage device and tracking log output processing method in multi-core environment

Abstract

The present invention is applicable to the field of storage technology, and provides a tracking log output processing method in a multi-core environment. The method includes: when the tracking log of each core is generated, the tracking log data is packaged, and time stamp information is added. Timestamp information is used to identify the sequence of trace log records; configure a lock-free circular cache structure and a static cache structure for each of the multiple cores, which are used to temporarily store the pointer of the trace log and temporarily store the data generated by each core. Tracking logs and corresponding timestamps; obtaining the timestamp information of the tracking logs according to the pointers of the tracking logs, and sorting the tracking logs of different static cache structures according to the timestamp information; when the timer is interrupted, the Trace log output processing. The present invention also correspondingly provides a multi-core storage device for realizing the above method. Thereby, the present invention can make the cache operation and the output operation run simultaneously, reduce the times of scheduling, and improve the utilization rate of resources.

Description

technical field [0001] The invention relates to the field of storage technology, in particular to a multi-core storage device and a tracking log output processing method in a multi-core environment. Background technique [0002] Trace log (trace log) is one of the common user processes in the storage system. In the prior art, a trace log buffer is generally applied for in the storage area, and the trace log is written into the buffer during program execution, and finally a task with the lowest priority is responsible for the output of the trace log. [0003] The above method has the following disadvantages: 1) When multiple cores write tracelog records to the trace log record buffer at the same time, they must first acquire the lock, so they need to wait for other Cores to release the lock; 2) In the Core where the task with the lowest priority is located When doing trace log record output, it is necessary to lock the trace log record buffer to avoid inconsistency in trace ...

AI Technical Summary

Problems solved by technology

[0003] The above method has the following disadvantages: 1) When multiple cores write tracelog records to the trace log record buffer at the same time, they must first acquire the lock, so they need to wait for other Cores to release the lock; 2) In the Core where the task with the lowest priority is located When doing trace log record output, it is necessary to lock the trace log record buffer to avoid inconsistency in trace log record management by locking, but this method will directly cause other cores to generate tracelog records at this time, then Must wait for the lowest priority task to release the lock

[0004] In summary, the existing technology obviously has inconvenience and defects in actual use, so it is necessary to improve

Images