executorlib/tests/unit/executor/test_flux_job_plot.py at 9050bd0bb8c1154816a3d1e62e8c5b59147f4a68 · pyiron/executorlib · 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
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
import os
import unittest
from time import sleep

from executorlib import FluxJobExecutor, FluxClusterExecutor
from executorlib.task_scheduler.interactive.dependency_plot import generate_nodes_and_edges_for_plotting
from executorlib.standalone.serialize import cloudpickle_register


try:
    import pygraphviz
    import flux.job
    from executorlib.task_scheduler.interactive.spawner_flux import FluxPythonSpawner

    skip_graphviz_flux_test = "FLUX_URI" not in os.environ
except ImportError:
    skip_graphviz_flux_test = True


def add_function(parameter_1, parameter_2):
    sleep(0.2)
    return parameter_1 + parameter_2


def generate_tasks(length):
    sleep(0.2)
    return range(length)


def calc_from_lst(lst, ind, parameter):
    sleep(0.2)
    return lst[ind] + parameter


def merge(lst):
    sleep(0.2)
    return sum(lst)


@unittest.skipIf(
    skip_graphviz_flux_test,
    "Either graphviz or flux are not installed, so the plot_dependency_graph tests are skipped.",
)
class TestFluxAllocationExecutorWithDependencies(unittest.TestCase):
    def test_executor_dependency_plot(self):
        with FluxJobExecutor(
            max_cores=1,
            plot_dependency_graph=True,
            block_allocation=False,
        ) as exe:
            cloudpickle_register(ind=1)
            future_1 = exe.submit(add_function, 1, parameter_2=2)
            future_2 = exe.submit(add_function, 1, parameter_2=future_1)
            self.assertTrue(future_1.done())
            self.assertTrue(future_2.done())
            self.assertEqual(len(exe._task_scheduler._future_hash_dict), 2)
            self.assertEqual(len(exe._task_scheduler._task_hash_dict), 2)
            nodes, edges = generate_nodes_and_edges_for_plotting(
                task_hash_dict=exe._task_scheduler._task_hash_dict,
                future_hash_inverse_dict={
                    v: k for k, v in exe._task_scheduler._future_hash_dict.items()
                },
            )
            self.assertEqual(len(nodes), 4)
            self.assertEqual(len(edges), 4)

    def test_many_to_one_plot(self):
        length = 5
        parameter = 1
        with FluxJobExecutor(
            max_cores=2,
            plot_dependency_graph=True,
            block_allocation=True,
        ) as exe:
            cloudpickle_register(ind=1)
            future_lst = exe.submit(
                generate_tasks,
                length=length,
                resource_dict={"cores": 1},
            )
            lst = []
            for i in range(length):
                lst.append(
                    exe.submit(
                        calc_from_lst,
                        lst=future_lst,
                        ind=i,
                        parameter=parameter,
                        resource_dict={"cores": 1},
                    )
                )
            future_sum = exe.submit(
                merge,
                lst=lst,
                resource_dict={"cores": 1},
            )
            self.assertTrue(future_lst.done())
            for l in lst:
                self.assertTrue(l.done())
            self.assertTrue(future_sum.done())
            self.assertEqual(len(exe._task_scheduler._future_hash_dict), 7)
            self.assertEqual(len(exe._task_scheduler._task_hash_dict), 7)
            nodes, edges = generate_nodes_and_edges_for_plotting(
                task_hash_dict=exe._task_scheduler._task_hash_dict,
                future_hash_inverse_dict={
                    v: k for k, v in exe._task_scheduler._future_hash_dict.items()
                },
            )
            self.assertEqual(len(nodes), 14)
            self.assertEqual(len(edges), 22)


@unittest.skipIf(
    skip_graphviz_flux_test,
    "Either graphviz or flux are not installed, so the plot_dependency_graph tests are skipped.",
)
class TestFluxSubmissionExecutorWithDependencies(unittest.TestCase):
    def test_executor_dependency_plot(self):
        with FluxClusterExecutor(
            plot_dependency_graph=True,
        ) as exe:
            cloudpickle_register(ind=1)
            future_1 = exe.submit(add_function, 1, parameter_2=2)
            future_2 = exe.submit(add_function, 1, parameter_2=future_1)
            self.assertTrue(future_1.done())
            self.assertTrue(future_2.done())
            self.assertEqual(len(exe._task_scheduler._future_hash_dict), 2)
            self.assertEqual(len(exe._task_scheduler._task_hash_dict), 2)
            nodes, edges = generate_nodes_and_edges_for_plotting(
                task_hash_dict=exe._task_scheduler._task_hash_dict,
                future_hash_inverse_dict={
                    v: k for k, v in exe._task_scheduler._future_hash_dict.items()
                },
            )
            self.assertEqual(len(nodes), 4)
            self.assertEqual(len(edges), 4)

    def test_many_to_one_plot(self):
        length = 5
        parameter = 1
        with FluxClusterExecutor(
            plot_dependency_graph=True,
        ) as exe:
            cloudpickle_register(ind=1)
            future_lst = exe.submit(
                generate_tasks,
                length=length,
                resource_dict={"cores": 1},
            )
            lst = []
            for i in range(length):
                lst.append(
                    exe.submit(
                        calc_from_lst,
                        lst=future_lst,
                        ind=i,
                        parameter=parameter,
                        resource_dict={"cores": 1},
                    )
                )
            future_sum = exe.submit(
                merge,
                lst=lst,
                resource_dict={"cores": 1},
            )
            self.assertTrue(future_lst.done())
            for l in lst:
                self.assertTrue(l.done())
            self.assertTrue(future_sum.done())
            self.assertEqual(len(exe._task_scheduler._future_hash_dict), 7)
            self.assertEqual(len(exe._task_scheduler._task_hash_dict), 7)
            nodes, edges = generate_nodes_and_edges_for_plotting(
                task_hash_dict=exe._task_scheduler._task_hash_dict,
                future_hash_inverse_dict={
                    v: k for k, v in exe._task_scheduler._future_hash_dict.items()
                },
            )
            self.assertEqual(len(nodes), 14)
            self.assertEqual(len(edges), 22)