Merge pull request #332 from 4dn-dcic/cost_estimate_force

Add --force to the cost_estimate command

Author: SooLee

docs/api.rst

@@ -586,3 +586,5 @@ Retrieve a cost estimate for a specific job. This will be available as soon as t
  update_tsv                         This flag specifies wether to update the cost in the tsv file that
                                     stores metrics information on the S3 bucket
 
+force                               Return the estimate, even if the actual cost is available
+

docs/commands.rst

@@ -591,4 +591,6 @@ This function requires a (deployed) Tibanna version >=1.0.6.
  -u|--update-tsv                     Update with the cost the tsv file that stores metrics
                                      information on the S3 bucket
 
+ -f|--force                          Return the estimate, even if the actual cost is available
+
 

docs/execution_json.rst

@@ -556,7 +556,7 @@ The ``config`` field describes execution configuration.
     - optional (default: unset)
 
 :ebs_throughput:
-    - Provisioned throughput of the gp3 type EBS (MiB/s). Must be an integer betweem 125 and 1000.
+    - Provisioned throughput of the gp3 type EBS (MiB/s). Must be an integer between 125 and 1000.
     - optional (default: unset)
 
 :ebs_type:

tibanna/__main__.py

@@ -310,7 +310,10 @@ def args(self):
                   'help': "job id of the specific job to log (alternative to --exec-arn/-e)"},
                  {'flag': ["-u", "--update-tsv"],
                   'help': "update estimated cost in the metric tsv file on S3",
-                  'action': "store_true"}],
+                  'action': "store_true"},
+                  {'flag': ["-f", "--force"],
+                  'action': "store_true",
+                  'help': "returns the estimate, even if the actual cost is available"}],
             'cleanup':
                 [{'flag': ["-g", "--usergroup"],
                   'help': "Tibanna usergroup that shares the permission to access buckets and run jobs"},
@@ -475,9 +478,9 @@ def cost(job_id, sfn=TIBANNA_DEFAULT_STEP_FUNCTION_NAME, update_tsv=False):
     """print out cost of a specific job"""
     print(API().cost(job_id=job_id, sfn=sfn, update_tsv=update_tsv))
 
-def cost_estimate(job_id, update_tsv=False):
+def cost_estimate(job_id, update_tsv=False, force=False):
     """print out estimated cost of a specific job"""
-    cost_estimate, cost_estimate_type = API().cost_estimate(job_id=job_id, update_tsv=update_tsv)
+    cost_estimate, cost_estimate_type = API().cost_estimate(job_id=job_id, update_tsv=update_tsv, force=force)
     print(f'{cost_estimate} ({cost_estimate_type})')
 
 

tibanna/core.py

@@ -1046,7 +1046,7 @@ def plot_metrics(self, job_id, sfn=None, directory='.', open_browser=True, force
         if open_browser:
             webbrowser.open(METRICS_URL(log_bucket, job_id))
 
-    def cost_estimate(self, job_id, update_tsv=False):
+    def cost_estimate(self, job_id, update_tsv=False, force=False):
         postrunjsonstr = self.log(job_id=job_id, postrunjson=True)
         if not postrunjsonstr:
             logger.info("Cost estimation error: postrunjson not found")
@@ -1055,12 +1055,13 @@ def cost_estimate(self, job_id, update_tsv=False):
         postrunjson = AwsemPostRunJson(**postrunjsonobj)
         log_bucket = postrunjson.config.log_bucket
 
-        # We return the real cost, if it is availble, but don't automatically update the Cost row in the tsv
-        precise_cost = self.cost(job_id, update_tsv=False)
-        if(precise_cost and precise_cost > 0.0):
-            if update_tsv:
-                update_cost_estimate_in_tsv(log_bucket, job_id, precise_cost, cost_estimate_type="actual cost")
-            return precise_cost, "actual cost"
+        # We return the real cost, if it is available, but don't automatically update the Cost row in the tsv
+        if not force:
+            precise_cost = self.cost(job_id, update_tsv=False)
+            if(precise_cost and precise_cost > 0.0):
+                if update_tsv:
+                    update_cost_estimate_in_tsv(log_bucket, job_id, precise_cost, cost_estimate_type="actual cost")
+                return precise_cost, "actual cost"
 
         # awsf_image was added in 1.0.0. We use that to get the correct ebs root type
         ebs_root_type = 'gp3' if 'awsf_image' in postrunjsonobj['config'] else 'gp2'

tibanna/pricing_utils.py

@@ -243,7 +243,6 @@ def get_cost_estimate(postrunjson, ebs_root_type = "gp3", aws_price_overwrite =
 
                 free_tier = 125
                 ebs_throughput_cost = ebs_throughput_price * max(cfg.ebs_throughput - free_tier, 0) * job_duration / (24.0*30.0)
-                print(ebs_throughput_cost)
                 estimated_cost = estimated_cost + ebs_throughput_cost
 
         else: