Method and device for testing omap operation performance and medium

Abstract

The invention discloses an omap operation performance testing method and device and a medium, and relates to the field of distributed storage. Test parameters are initialized, wherein the test parameters at least comprise running time and concurrency; the length of the operation queue is equal to the concurrency number; an omap request is generated, the omap request comprises an omap write request or an omap read request, the number of the omap requests is equal to the concurrent number, and each omap request corresponds to the slot position of the operation queue one by one; asynchronous processing is conducted on the omap request through the operation queue, whether the processing process meets the preset condition or not is judged, and if yes, test statistical data are output. Therefore, according to the scheme, the write operation and the read operation of the omap operation in the distributed storage system can be tested respectively, and due to the fact that the concurrency number is set and an asynchronous processing mode is adopted, the pressure intensity for controlling the test can be sent in parallel, so that the limit value of the omap operation performance is obtained, and the method is very reliable.

Description

technical field [0001] The present application relates to the field of distributed storage, and in particular, to a method, device and medium for testing omap operation performance. Background technique [0002] omap refers to object map, which is an object used to store key-value key-value pair map data. In distributed storage, omap objects play a very important role. In the file and object services provided by distributed storage, the performance of omap directly affects the performance of cluster storage. In the file service, metadata such as directory indexes are stored in omap objects, such as recording information such as files corresponding to a directory; in addition, the free directory objects that are asynchronously deleted in the file system are also omap objects, and the files to be deleted are first put into Free directory, and then slowly clean up in the background. There are also garbage collection objects similar to free directory objects in the object stor...

AI Technical Summary

Problems solved by technology

[0003] Generally, each local storage engine has its own performance benchmarking tool and provides its own performance benchmarking results to the outside, but after the local storage engine is integrated with the storage system, there is no implementation of the local database interface for the storage system. performance benchmarking

If the performance benchmark test is performed directly in the form of file or object service, the performance indicators are not measurable and not universal, and the code call stack is deepened, resulting in performance bottlenecks; Measured limit value

Images