Home • Explore • Blog • CI
☰



kubernetes
3rd chunk of `content/en/blog/_posts/2017-12-00-Paddle-Paddle-Fluid-Elastic-Learning.md`
9ebd4c1cbe4b3008f61f67eed97e9466ff352859c9c0359f000000010000047a
![](https://1.bp.blogspot.com/-sp_sVZvhMbU/WiYgXMLQKuI/AAAAAAAAAIM/uc_3iT9BZmAtQGiGGSErgueHK71uWMBCACEwYBhgL/s640/figure-1.png)](https://1.bp.blogspot.com/-sp_sVZvhMbU/WiYgXMLQKuI/AAAAAAAAAIM/uc_3iT9BZmAtQGiGGSErgueHK71uWMBCACEwYBhgL/s1600/figure-1.png) |
| _Figure 1. Fluid EDL evenly distributes resource among jobs._  
 |


In the second test, each experiment ran 400 Nginx pods, which has higher priority than the six PaddlePaddle jobs. Initially, each PaddlePaddle job had 15 trainers and 10 parameter servers. We killed 100 Nginx pods every 90 seconds until 100 left, and then we started to increase the number of Nginx jobs by 100 every 90 seconds. The upper part of Figure 2 shows this process. The middle of the diagram shows that Fluid EDL automatically started some PaddlePaddle processes by decreasing Nginx pods, and killed PaddlePaddle processes by increasing Nginx pods later on. As a result, the cluster maintains around 90% utilization as shown in the bottom of the figure. When Fluid EDL was turned off, there were no PaddlePaddle processes autoincrement, and the utilization fluctuated with the varying number of Nginx pods.
Title: Fluid EDL and Resource Prioritization with Nginx Jobs
Summary
In the second test, the performance of Fluid EDL was assessed in a cluster with Nginx pods (higher priority) and PaddlePaddle jobs. Nginx pods were gradually reduced and then increased. Fluid EDL dynamically adjusted the number of PaddlePaddle processes based on the availability of resources due to Nginx pod fluctuation, maintaining approximately 90% cluster utilization. Without Fluid EDL, PaddlePaddle processes did not automatically adjust, causing utilization to fluctuate with the number of Nginx pods.