def test_purging_output_cleaned_after_ok_run(self):
history_id, job_state, outputs = self._setup_running_two_output_job(10)
# Pretty much right away after the job is running, these paths should be populated -
# if they are grab them and make sure they are deleted at the end of the job.
dataset_1 = self._get_history_item_as_admin(history_id,
outputs[0]["id"])
dataset_2 = self._get_history_item_as_admin(history_id,
outputs[1]["id"])
if "file_name" in dataset_1:
output_dataset_paths = [
dataset_1["file_name"], dataset_2["file_name"]
]
# This may or may not exist depending on if the test is local or not.
output_dataset_paths_exist = os.path.exists(
output_dataset_paths[0])
else:
output_dataset_paths = []
output_dataset_paths_exist = False
if not output_dataset_paths_exist:
# Given this Galaxy configuration - there is nothing more to be tested here.
# Consider throwing a skip instead.
return
# Purge one of the two outputs and wait for the job to complete.
self._raw_update_history_item(history_id, outputs[0]["id"],
{"purged": True})
wait_on_state(job_state, assert_ok=True)
if output_dataset_paths_exist:
time.sleep(.5)
# Make sure the non-purged dataset is on disk and the purged one is not.
assert os.path.exists(output_dataset_paths[1])
assert not os.path.exists(output_dataset_paths[0])
def test_purging_output_cleaned_after_ok_run(self):
history_id, job_state, outputs = self._setup_running_two_output_job(10)
# Pretty much right away after the job is running, these paths should be populated -
# if they are grab them and make sure they are deleted at the end of the job.
dataset_1 = self._get_history_item_as_admin(history_id, outputs[0]["id"])
dataset_2 = self._get_history_item_as_admin(history_id, outputs[1]["id"])
if "file_name" in dataset_1:
output_dataset_paths = [dataset_1["file_name"], dataset_2["file_name"]]
# This may or may not exist depending on if the test is local or not.
output_dataset_paths_exist = os.path.exists(output_dataset_paths[0])
else:
output_dataset_paths = []
output_dataset_paths_exist = False
if not output_dataset_paths_exist:
# Given this Galaxy configuration - there is nothing more to be tested here.
# Consider throwing a skip instead.
return
# Purge one of the two outputs and wait for the job to complete.
self._raw_update_history_item(history_id, outputs[0]["id"], {"purged": True})
wait_on_state(job_state, assert_ok=True)
if output_dataset_paths_exist:
time.sleep(.5)
# Make sure the non-purged dataset is on disk and the purged one is not.
assert os.path.exists(output_dataset_paths[1])
assert not os.path.exists(output_dataset_paths[0])
def _setup_running_two_output_job(self, sleep_time):
history_id = self.dataset_populator.new_history()
payload = self.dataset_populator.run_tool_payload(
tool_id='create_2',
inputs=dict(sleep_time=sleep_time, ),
history_id=history_id,
)
run_response = self._post("tools", data=payload).json()
outputs = run_response["outputs"]
jobs = run_response["jobs"]
assert len(outputs) == 2
assert len(jobs) == 1
def job_state():
jobs_response = self._get("jobs/%s" % jobs[0]["id"])
return jobs_response
# Give job some time to get up and running.
time.sleep(2)
running_state = wait_on_state(job_state,
skip_states=["queued", "new"],
assert_ok=False,
timeout=15)
assert running_state == "running", running_state
def job_state():
jobs_response = self._get("jobs/%s" % jobs[0]["id"])
return jobs_response
return history_id, job_state, outputs
def _setup_running_two_output_job(self, sleep_time):
history_id = self.dataset_populator.new_history()
payload = self.dataset_populator.run_tool_payload(
tool_id='create_2',
inputs=dict(
sleep_time=sleep_time,
),
history_id=history_id,
)
run_response = self._post("tools", data=payload).json()
outputs = run_response["outputs"]
jobs = run_response["jobs"]
assert len(outputs) == 2
assert len(jobs) == 1
def job_state():
jobs_response = self._get("jobs/%s" % jobs[0]["id"])
return jobs_response
# Give job some time to get up and running.
time.sleep(2)
running_state = wait_on_state(job_state, skip_states=["queued", "new"], assert_ok=False, timeout=15)
assert running_state == "running", running_state
def job_state():
jobs_response = self._get("jobs/%s" % jobs[0]["id"])
return jobs_response
return history_id, job_state, outputs
def test_purging_output_keep_running_until_all_purged( self ):
history_id, job_state, outputs = self._setup_running_two_output_job( 120 )
# Pretty much right away after the job is running, these paths should be populated -
# if they are grab them and make sure they are deleted at the end of the job.
dataset_1 = self._get_history_item_as_admin( history_id, outputs[0]["id"] )
dataset_2 = self._get_history_item_as_admin( history_id, outputs[1]["id"] )
if "file_name" in dataset_1:
output_dataset_paths = [ dataset_1[ "file_name" ], dataset_2[ "file_name" ] ]
# This may or may not exist depending on if the test is local or not.
output_dataset_paths_exist = os.path.exists( output_dataset_paths[ 0 ] )
else:
output_dataset_paths = []
output_dataset_paths_exist = False
current_state = job_state().json()["state"]
assert current_state == "running", current_state
# Purge one of the two outputs and make sure the job is still running.
self._raw_update_history_item( history_id, outputs[0]["id"], {"purged": True} )
time.sleep( 1 )
current_state = job_state().json()["state"]
assert current_state == "running", current_state
# Purge the second output and make sure the job is cancelled.
self._raw_update_history_item( history_id, outputs[1]["id"], {"purged": True} )
final_state = wait_on_state( job_state, assert_ok=False, timeout=15 )
assert final_state in ["deleted_new", "deleted"], final_state
def paths_deleted():
if not os.path.exists( output_dataset_paths[ 0 ] ) and not os.path.exists( output_dataset_paths[ 1 ] ):
return True
if output_dataset_paths_exist:
wait_on(paths_deleted, "path deletion")
def test_create_dataset( self ):
library = self.library_populator.new_private_library( "ForCreateDatasets" )
payload, files = self.library_populator.create_dataset_request( library, file_type="txt", contents="create_test" )
create_response = self._post( "libraries/%s/contents" % library[ "id" ], payload, files=files )
self._assert_status_code_is( create_response, 200 )
library_datasets = create_response.json()
assert len( library_datasets ) == 1
library_dataset = library_datasets[ 0 ]
def show():
return self._get( "libraries/%s/contents/%s" % ( library[ "id" ], library_dataset[ "id" ] ) )
wait_on_state( show, assert_ok=True )
library_dataset = show().json()
self._assert_has_keys( library_dataset, "peek", "data_type" )
assert library_dataset[ "peek" ].find("create_test") >= 0
assert library_dataset[ "file_ext" ] == "txt", library_dataset[ "file_ext" ]
def test_deleting_output_keep_running_until_all_deleted( self ):
history_id, job_state, outputs = self._setup_running_two_output_job( 120 )
# Delete one of the two outputs and make sure the job is still running.
self._raw_update_history_item( history_id, outputs[0]["id"], {"deleted": True} )
time.sleep( 1 )
state = job_state().json()["state"]
assert state == "running", state
# Delete the second output and make sure the job is cancelled.
self._raw_update_history_item( history_id, outputs[1]["id"], {"deleted": True} )
final_state = wait_on_state( job_state, assert_ok=False, timeout=15 )
assert final_state in ["deleted_new", "deleted"], final_state
def test_create_dataset(self):
library = self.library_populator.new_private_library(
"ForCreateDatasets")
payload, files = self.library_populator.create_dataset_request(
library, file_type="txt", contents="create_test")
create_response = self._post("libraries/%s/contents" % library["id"],
payload,
files=files)
self._assert_status_code_is(create_response, 200)
library_datasets = create_response.json()
assert len(library_datasets) == 1
library_dataset = library_datasets[0]
def show():
return self._get("libraries/%s/contents/%s" %
(library["id"], library_dataset["id"]))
wait_on_state(show, assert_ok=True)
library_dataset = show().json()
self._assert_has_keys(library_dataset, "peek", "data_type")
assert library_dataset["peek"].find("create_test") >= 0
assert library_dataset["file_ext"] == "txt", library_dataset[
"file_ext"]
