Advances, Systems and Applications
From: Task grouping and optimized deep learning based VM sizing for hosting containers as a service
Number of Iteration | Deep-ConvLSTM_FPO with Population Size 5 | Deep-ConvLSTM_FPO with Population Size 10 | Deep-ConvLSTM_FPO with Population Size 15 | Deep-ConvLSTM_FPO with Population Size 20 |
---|---|---|---|---|
Resource Utilization | ||||
5 | 0.0987 | 0.0958 | 0.0925 | 0.0908 |
10 | 0.1025 | 0.0995 | 0.0976 | 0.0943 |
15 | 0.1025 | 0.0987 | 0.0965 | 0.0937 |
20 | 0.1041 | 0.0999 | 0.0975 | 0.0936 |
Response Time | ||||
5 | 650 | 649 | 646 | 643 |
10 | 527 | 525 | 522 | 520 |
15 | 511 | 508 | 503 | 501 |
20 | 482 | 478 | 475 | 472 |
Task Rejection Rate | ||||
5 | 0.3254 | 0.3168 | 0.3088 | 0.2987 |
10 | 0.3303 | 0.3268 | 0.3168 | 0.3077 |
15 | 0.3401 | 0.3368 | 0.3268 | 0.3168 |
20 | 0.3387 | 0.3258 | 0.3168 | 0.3088 |
Makespan | ||||
5 | 0.6088 | 0.5968 | 0.5765 | 0.5645 |
10 | 0.6257 | 0.6146 | 0.6079 | 0.5975 |
15 | 0.6579 | 0.6466 | 0.6257 | 0.6146 |
20 | 0.6755 | 0.6644 | 0.6457 | 0.6368 |