Nitro-VLDB25/lambda_scale_test.py at master · IntelliSys-Lab/Nitro-VLDB25 · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
import numpy as np
import collections
import logging
import ray
from ray.rllib.policy.sample_batch import DEFAULT_POLICY_ID
from env import Environment
import config
import utils


def lambda_scale_test(
    scheduler_name,
    num_rollout_workers,
    algo_name,
    env_name,
):
    # Set up environment and load checkpoint
    env = Environment(
        scheduler_name=scheduler_name,
        algo_name=algo_name,
        env_name=env_name,
        target_reward=config.envs[env_name]["max_reward"],
        budget=config.envs[env_name]["budget"],
        stop_min_round=config.stop_min_round,
        stop_max_round=config.stop_max_round,
        stop_num_results=config.stop_num_results,
        stop_cv=config.stop_cv,
        stop_grace_period=config.stop_grace_period,
        is_serverless=True,
    )

    # Start training
    state, mask, info = env.reset()

    payload = {
        "redis_host": config.redis_host,
        "redis_port": config.redis_port,
        "redis_password": config.redis_password,
        "algo_name": algo_name,
        "env_name": env_name,
        "num_envs_per_worker": 1,
        "rollout_fragment_length": config.envs[env_name]['rollout_fragment_length'],
    }

    invoke_overhead, query_overhead = env.scale_test(
        num_rollout_workers=num_rollout_workers,
        payload=payload,
    )

    print("")
    print("******************")
    print("******************")
    print("******************")
    print("")
    print("Running {}, algo {}, env {}".format(scheduler_name, algo_name, env_name))
    print("invoke_overhead: {}".format(invoke_overhead))
    print("query_overhead: {}".format(query_overhead))

    env.stop_trainer()

    return invoke_overhead, query_overhead


if __name__ == '__main__':
    scheduler_name = "lambda_scale_test"
    num_rollout_workers_list = [1000, 900, 800, 700, 600, 500, 400, 300, 200, 100, 1]
    csv_invoke_overhead = [
        [
            "Hopper",
            "Humanoid",
            "Walker2d",
            "Gravitar",
            "SpaceInvaders",
            "Qbert",
        ]
    ]
    csv_query_overhead = [
        [
            "Hopper",
            "Humanoid",
            "Walker2d",
            "Gravitar",
            "SpaceInvaders",
            "Qbert",
        ]
    ]

    print("")
    print("**********")
    print("**********")
    print("**********")
    print("")
    ray.init(
        log_to_driver=False,
        configure_logging=True,
        logging_level=logging.ERROR
    )

    for num_rollout_workers in num_rollout_workers_list:
        for algo_name in config.algos:
            csv_env_invoke_overhead = []
            csv_env_query_overhead = []

            for env_name in config.envs.keys():
                invoke_overhead, query_overhead = lambda_scale_test(
                    scheduler_name=scheduler_name,
                    num_rollout_workers=num_rollout_workers,
                    algo_name=algo_name,
                    env_name=env_name,
                )
                csv_env_invoke_overhead.append(invoke_overhead)
                csv_env_query_overhead.append(query_overhead)

            csv_invoke_overhead.append(csv_env_invoke_overhead)
            csv_query_overhead.append(csv_env_query_overhead)

    utils.export_csv(
        scheduler_name=scheduler_name,
        env_name="",
        algo_name="",
        csv_name="invoke_overhead",
        csv_file=csv_invoke_overhead
    )
    utils.export_csv(
        scheduler_name=scheduler_name,
        env_name="",
        algo_name="",
        csv_name="query_overhead",
        csv_file=csv_query_overhead
    )


    ray.shutdown()
    print("")
    print("**********")
    print("**********")
    print("**********")