changed tests to check for specific usage values

also changed the values to be more recognizable

Author: VictorVerhaert

openeo/rest/_testing.py

@@ -231,11 +231,11 @@ def _handle_get_job(self, request, context):
             "status": self.batch_jobs[job_id]["status"],
         }
         if self.batch_jobs[job_id]["status"] == "finished":  # HACK some realistic values for a small job
-            result["costs"] = 4
+            result["costs"] = 123
             result["usage"] = {
-                "cpu": {"unit": "cpu-seconds", "value": 30.0},
-                "duration": {"unit": "seconds", "value": 55},
-                "memory": {"unit": "mb-seconds", "value": 150000.0},
+                "cpu": {"unit": "cpu-seconds", "value": 1234.5},
+                "memory": {"unit": "mb-seconds", "value": 34567.89},
+                "duration": {"unit": "seconds", "value": 2345},
             }
         return result
 

tests/extra/test_job_management.py

@@ -18,6 +18,7 @@
 #   httpretty avoids this specific problem because it mocks at the socket level,
 #   But I would rather not have two dependencies with almost the same goal.
 import httpretty
+import numpy as np
 import pandas
 import pandas as pd
 import pytest
@@ -166,16 +167,17 @@ def test_basic(self, tmp_path, job_manager, job_manager_root_dir, sleep_mock):
             }
         )
 
-        assert [(r.id, r.status, r.backend_name) for r in pd.read_csv(job_db_path).itertuples()] == [
-            ("job-2018", "finished", "foo"),
-            ("job-2019", "finished", "foo"),
-            ("job-2020", "finished", "bar"),
-            ("job-2021", "finished", "bar"),
-            ("job-2022", "finished", "foo"),
+        assert [
+            (r.id, r.status, r.backend_name, r.cpu, r.memory, r.duration, r.costs)
+            for r in pd.read_csv(job_db_path).itertuples()
+        ] == [
+            ("job-2018", "finished", "foo", "1234.5 cpu-seconds", "34567.89 mb-seconds", "2345 seconds", 123),
+            ("job-2019", "finished", "foo", "1234.5 cpu-seconds", "34567.89 mb-seconds", "2345 seconds", 123),
+            ("job-2020", "finished", "bar", "1234.5 cpu-seconds", "34567.89 mb-seconds", "2345 seconds", 123),
+            ("job-2021", "finished", "bar", "1234.5 cpu-seconds", "34567.89 mb-seconds", "2345 seconds", 123),
+            ("job-2022", "finished", "foo", "1234.5 cpu-seconds", "34567.89 mb-seconds", "2345 seconds", 123),
         ]
 
-        assert not pd.read_csv(job_db_path)[["cpu", "memory", "duration", "costs"]].isnull().any().any()
-
         # Check downloaded results and metadata.
         assert set(p.relative_to(job_manager_root_dir) for p in job_manager_root_dir.glob("**/*.*")) == {
             Path(f"job_{job_id}") / filename
@@ -206,7 +208,10 @@ def test_db_class(self, tmp_path, job_manager, job_manager_root_dir, sleep_mock,
         assert len(result) == 5
         assert set(result.status) == {"finished"}
         assert set(result.backend_name) == {"foo", "bar"}
-        assert not result[["cpu", "memory", "duration", "costs"]].isnull().any().any()
+        assert set(result.cpu) == {"1234.5 cpu-seconds"}
+        assert set(result.memory) == {"34567.89 mb-seconds"}
+        assert set(result.duration) == {"2345 seconds"}
+        assert set(result.costs) == {123}
 
     @pytest.mark.parametrize(
         ["filename", "expected_db_class"],
@@ -257,16 +262,17 @@ def test_basic_threading(self, tmp_path, job_manager, job_manager_root_dir, slee
         # TODO #645 how to collect stats with the threaded run_job?
         assert sleep_mock.call_count > 10
 
-        assert [(r.id, r.status, r.backend_name) for r in pd.read_csv(job_db_path).itertuples()] == [
-            ("job-2018", "finished", "foo"),
-            ("job-2019", "finished", "foo"),
-            ("job-2020", "finished", "bar"),
-            ("job-2021", "finished", "bar"),
-            ("job-2022", "finished", "foo"),
+        assert [
+            (r.id, r.status, r.backend_name, r.cpu, r.memory, r.duration, r.costs)
+            for r in pd.read_csv(job_db_path).itertuples()
+        ] == [
+            ("job-2018", "finished", "foo", "1234.5 cpu-seconds", "34567.89 mb-seconds", "2345 seconds", 123),
+            ("job-2019", "finished", "foo", "1234.5 cpu-seconds", "34567.89 mb-seconds", "2345 seconds", 123),
+            ("job-2020", "finished", "bar", "1234.5 cpu-seconds", "34567.89 mb-seconds", "2345 seconds", 123),
+            ("job-2021", "finished", "bar", "1234.5 cpu-seconds", "34567.89 mb-seconds", "2345 seconds", 123),
+            ("job-2022", "finished", "foo", "1234.5 cpu-seconds", "34567.89 mb-seconds", "2345 seconds", 123),
         ]
 
-        assert not pd.read_csv(job_db_path)[["cpu", "memory", "duration", "costs"]].isnull().any().any()
-
         # Check downloaded results and metadata.
         assert set(p.relative_to(job_manager_root_dir) for p in job_manager_root_dir.glob("**/*.*")) == {
             Path(f"job_{job_id}") / filename
@@ -339,15 +345,16 @@ def start_worker_thread():
         )
 
         # Also check that we got sensible end results in the job db.
-        result = pd.read_csv(job_db_path)
-        assert [(r.id, r.status, r.backend_name) for r in result.itertuples()] == [
-            ("job-2018", "finished", "foo"),
-            ("job-2019", "finished", "foo"),
-            ("job-2020", "finished", "bar"),
-            ("job-2021", "finished", "bar"),
-            ("job-2022", "error", "foo"),
+        results = pd.read_csv(job_db_path).replace({np.nan: None})  # np.nan's are replaced by None for easy comparison
+        assert [
+            (r.id, r.status, r.backend_name, r.cpu, r.memory, r.duration, r.costs) for r in results.itertuples()
+        ] == [
+            ("job-2018", "finished", "foo", "1234.5 cpu-seconds", "34567.89 mb-seconds", "2345 seconds", 123),
+            ("job-2019", "finished", "foo", "1234.5 cpu-seconds", "34567.89 mb-seconds", "2345 seconds", 123),
+            ("job-2020", "finished", "bar", "1234.5 cpu-seconds", "34567.89 mb-seconds", "2345 seconds", 123),
+            ("job-2021", "finished", "bar", "1234.5 cpu-seconds", "34567.89 mb-seconds", "2345 seconds", 123),
+            ("job-2022", "error", "foo", None, None, None, None),
         ]
-        assert not result[result["status"] == "finished"][["cpu", "memory", "duration", "costs"]].isnull().any().any()
 
         # Check downloaded results and metadata.
         assert set(p.relative_to(job_manager_root_dir) for p in job_manager_root_dir.glob("**/*.*")) == {