How to test the concurrency capability of streaming media server?

Generally, in video transmission or distribution, in addition to paying attention to the stability of the system, everyone is more concerned about the distribution or concurrency capabilities of streaming media. In the TSINGSEE Qingxi video system, concurrently refers to the number of clients accessing the video, to put it bluntly, the number of computers or the number of people playing the video. During the deployment of some projects, users may have high concurrency requirements. This article will share with you how to test the concurrency capability of streaming media.

Test content: Test the maximum distribution capacity of the media server (multiple users watch a camera)

Preconditions:

  • The system boots and connects a camera to the system platform
  • Write an application program to simulate users watching a video, and add 50 users to watch a video every 60s

Test steps:

  • Use the application program on the client side to perform video viewing operations, adding 50 users every 60s, for a total of 500 users
  • Record the occupancy, memory occupancy, and bandwidth of the media server for every 50 additional users
  • Turn on a machine, monitor the video, and observe whether the video is playing normally
  • Record the maximum number of clients supported by the media server when it runs stably (to ensure that the client can watch the video normally)
  • Cycle through steps 1-4 and take the average

Test content: Test the concurrent processing capability of the media server (multiple users watch a camera at the same time)

Preconditions:

  • System boots, a camera is connected to the system platform
  • Write an application program to simulate users watching videos, and initiate a request for 1000 users to watch videos at a time

Test steps:

  • Turn on a machine and monitor a video to ensure the normal playback of the video
  • Use the application program on the client side to perform video viewing operations, and initiate 1000 viewing video requests at a time
  • Record the occupancy, memory occupancy, bandwidth, and response time of the media server when it is running stably (to ensure that the client can watch the video normally)
  • Cycle through steps 1-3 and take the average

Test content: Test the stability of the media server

Preconditions:

  • System startup, camera connected to the system platform
  • Write an application program to simulate users watching videos, and add 50 users to watch a video every 60s, increasing the number of users to 500 in total

Test steps:

  • Turn on a computer and monitor a video to ensure the normal playback of the video
  • Use the application program on the client side to perform video viewing operations, adding 50 users every 60s, a total of 500 users, making the platform run at high load
  • Record the stable running time of the platform (to ensure that the client can watch the video normally), server occupancy, memory occupancy, and network bandwidth

Test content: test the maximum number of recordings supported by the system within the bandwidth of 1000M

Preconditions:

  • System startup, camera connected to the system platform
  • The server is deployed under 1000M bandwidth

Test steps:

  • Set up a period of 2 hours central storage video for 100 terminals (D1 format) at the same time
  • After the recording time starts, check whether the 100-segment recording starts recording, and whether the recording stops when the recording end time is up
  • Record server consumption, memory usage and bandwidth

Guess you like

Origin blog.csdn.net/TsingSee/article/details/120485564#comments_25830328