Performance test method of NFS and performance test device of NFS

Abstract

The invention provides a performance test method of an NFS (Network File System) and a performance test device of the NFS. The test method comprises the following steps of starting an automatic deployment script, wherein a test script, a configuration file and a test tool are deployed to other NFS clients; starting all of the NFS clients, wherein all of the NFS clients synchronously performs test on a service end through the test tool, and the test result is stored in a corresponding test log; and under the test completion condition, collecting and naming test logs of all of the clients through an automatic filing script. Through a multi-client automatic test technology of the NFS service response time delay in the cluster NAS, a plurality of client nodes can synchronously perform the NFS performance test; the manual interference is not needed; and meanwhile, a test result automatic filing technology is used, so that the logs generated by the plurality of client nodes can be timely collected and sorted to a main starting client, and the later-stage unified comparison analysis is convenient.

Description

technical field [0001] The invention relates to the field of performance testing, in particular to a performance testing method and device of a network file system NFS. Background technique [0002] The storage system is an important part of the computer system. In the current era of big data, higher requirements are placed on the performance of the storage system. Clustered NAS (NetworkAttachedStorage, Network Attached Storage) is one of the essential technologies in the storage system. It is a loosely coupled collection of computing nodes composed of multiple nodes, which work together to provide high performance, high availability, and high load. Balanced NFS / CIFS (Network File System / Common Network File System) storage service. Testing the performance of NFS services provided by clustered NAS is one of the indispensable steps in the system development process. [0003] The existing test methods are mainly as follows: First, pre-organize a large amount of data on the cl...

AI Technical Summary

Problems solved by technology

[0004] The existing test methods mainly have the following problems: First, the test indicators are relatively single, and more application situations are not considered. For example, in the media industry mainly for video and audio data, the more important indicator of the NFS service of the clustered NAS is the response The maximum delay, where the maximum delay refers to the maximum value of the time taken by the NFS client to send the request to receive the response data sent by the NFS server in the cluster NAS

Secondly, in a large storage system, there may be a large number of clients that mount the clustered NAS through NFS. Each test needs to deploy the same configuration on the client and start the operation together. Such a repetitive process will waste a lot of labor costs.

Finally, a large number of log records generated after each test need to be collected and sorted in a timely manner. Otherwise, after a period of time, under certain unpredictable factors (such as: system power failure, time out of sync), the same experiment of different clients Record data may not be unified together

[0005] For the problems in related technologies, no effective solutions have been proposed yet

Images