Merge branch '2411-check-copyright'

[arvados.git] / sdk / cwl / arvados_cwl / arvcontainer.py

diff --git a/sdk/cwl/arvados_cwl/arvcontainer.py b/sdk/cwl/arvados_cwl/arvcontainer.py
index 374f368128afe77cb97d10e11aab6a9ac52757bb..4ab65d9d8774708613787b3a694f64bf876004da 100644 (file)
--- a/sdk/cwl/arvados_cwl/arvcontainer.py
+++ b/sdk/cwl/arvados_cwl/arvcontainer.py
@@ -1,7 +1,14 @@
+# Copyright (C) The Arvados Authors. All rights reserved.
+#
+# SPDX-License-Identifier: Apache-2.0
+
 import logging
 import json
 import os
 import urllib
+import time
+import datetime
+import ciso8601
 
 import ruamel.yaml as yaml
 
@@ -181,9 +188,15 @@ class ArvadosContainer(object):
         container_request["output_ttl"] = self.output_ttl
         container_request["mounts"] = mounts
         container_request["runtime_constraints"] = runtime_constraints
-        container_request["use_existing"] = kwargs.get("enable_reuse", True)
         container_request["scheduling_parameters"] = scheduling_parameters
 
+        enable_reuse = kwargs.get("enable_reuse", True)
+        if enable_reuse:
+            reuse_req, _ = get_feature(self, "http://arvados.org/cwl#ReuseRequirement")
+            if reuse_req:
+                enable_reuse = reuse_req["enableReuse"]
+        container_request["use_existing"] = enable_reuse
+
         if kwargs.get("runnerjob", "").startswith("arvwf:"):
             wfuuid = kwargs["runnerjob"][6:kwargs["runnerjob"].index("#")]
             wfrecord = self.arvrunner.api.workflows().get(uuid=wfuuid).execute(num_retries=self.arvrunner.num_retries)
@@ -211,9 +224,6 @@ class ArvadosContainer(object):
     def done(self, record):
         outputs = {}
         try:
-            if self.output_ttl:
-                self.arvrunner.add_intermediate_output(record["output_uuid"])
-
             container = self.arvrunner.api.containers().get(
                 uuid=record["container_uuid"]
             ).execute(num_retries=self.arvrunner.num_retries)
@@ -239,6 +249,17 @@ class ArvadosContainer(object):
                                                            num_retries=self.arvrunner.num_retries)
                 done.logtail(logc, logger, "%s error log:" % self.arvrunner.label(self))
 
+            if record["output_uuid"]:
+                if self.arvrunner.trash_intermediate or self.arvrunner.intermediate_output_ttl:
+                    # Compute the trash time to avoid requesting the collection record.
+                    trash_at = ciso8601.parse_datetime_unaware(record["modified_at"]) + datetime.timedelta(0, self.arvrunner.intermediate_output_ttl)
+                    aftertime = " at %s" % trash_at.strftime("%Y-%m-%d %H:%M:%S UTC") if self.arvrunner.intermediate_output_ttl else ""
+                    orpart = ", or" if self.arvrunner.trash_intermediate and self.arvrunner.intermediate_output_ttl else ""
+                    oncomplete = " upon successful completion of the workflow" if self.arvrunner.trash_intermediate else ""
+                    logger.info("%s Intermediate output %s (%s) will be trashed%s%s%s." % (
+                        self.arvrunner.label(self), record["output_uuid"], container["output"], aftertime, orpart, oncomplete))
+                self.arvrunner.add_intermediate_output(record["output_uuid"])
+
             if container["output"]:
                 outputs = done.done_outputs(self, container, "/tmp", self.outdir, "/keep")
         except WorkflowException as e:
@@ -339,6 +360,9 @@ class RunnerContainer(Runner):
         if self.intermediate_output_ttl:
             command.append("--intermediate-output-ttl=%d" % self.intermediate_output_ttl)
 
+        if self.arvrunner.trash_intermediate:
+            command.append("--trash-intermediate")
+
         command.extend([workflowpath, "/var/lib/cwl/cwl.input.json"])
 
         container_req["command"] = command