20825: Fix git_info

[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 632e171b40220f90e57663bcbc363f25d5b86dac..487ab70909a95b5f3849474fae360935e6a2794d 100644 (file)
--- a/sdk/cwl/arvados_cwl/arvcontainer.py
+++ b/sdk/cwl/arvados_cwl/arvcontainer.py
@@ -367,11 +367,11 @@ class ArvadosContainer(JobBase):
                 logger.warning("%s API revision is %s, revision %s is required to support setting properties on output collections.",
                                self.arvrunner.label(self), self.arvrunner.api._rootDesc["revision"], "20220510")
 
-        ramMultiplier = [1]
+        ram_multiplier = [1]
 
         oom_retry_req, _ = self.get_requirement("http://arvados.org/cwl#OutOfMemoryRetry")
         if oom_retry_req and oom_retry_req.get('memoryRetryMultipler'):
-            ramMultiplier.append(oom_retry_req.get('memoryRetryMultipler'))
+            ram_multiplier.append(oom_retry_req.get('memoryRetryMultipler'))
 
         if runtimeContext.runnerjob.startswith("arvwf:"):
             wfuuid = runtimeContext.runnerjob[6:runtimeContext.runnerjob.index("#")]
@@ -386,12 +386,14 @@ class ArvadosContainer(JobBase):
         try:
             ram = runtime_constraints["ram"]
 
-            for i in ramMultiplier:
+            self.uuid = runtimeContext.submit_request_uuid
+
+            for i in ram_multiplier:
                 runtime_constraints["ram"] = ram * i
 
-                if runtimeContext.submit_request_uuid:
+                if self.uuid:
                     response = self.arvrunner.api.container_requests().update(
-                        uuid=runtimeContext.submit_request_uuid,
+                        uuid=self.uuid,
                         body=container_request,
                         **extra_submit_params
                     ).execute(num_retries=self.arvrunner.num_retries)
@@ -400,22 +402,21 @@ class ArvadosContainer(JobBase):
                         body=container_request,
                         **extra_submit_params
                     ).execute(num_retries=self.arvrunner.num_retries)
-                    runtimeContext.submit_request_uuid = response["uuid"]
+                    self.uuid = response["uuid"]
 
                 if response["container_uuid"] is not None:
                     break
 
             if response["container_uuid"] is None:
-                runtime_constraints["ram"] = ram * ramMultiplier[self.attempt_count]
+                runtime_constraints["ram"] = ram * ram_multiplier[self.attempt_count]
 
             container_request["state"] = "Committed"
             response = self.arvrunner.api.container_requests().update(
-                uuid=runtimeContext.submit_request_uuid,
+                uuid=self.uuid,
                 body=container_request,
                 **extra_submit_params
             ).execute(num_retries=self.arvrunner.num_retries)
 
-            self.uuid = response["uuid"]
             self.arvrunner.process_submitted(self)
             self.attempt_count += 1
 
@@ -449,7 +450,7 @@ class ArvadosContainer(JobBase):
         done.logtail(logc, callback, "", maxlen=1000)
 
         # Check allocation failure
-        oom_matches = oom_retry_req.get('memoryErrorRegex') or r'(bad_alloc|out ?of ?memory|container using over 9.% of memory)'
+        oom_matches = oom_retry_req.get('memoryErrorRegex') or r'(bad_alloc|out ?of ?memory|memory ?error|container using over 9.% of memory)'
         if re.search(oom_matches, loglines[0], re.IGNORECASE | re.MULTILINE):
             return True
 
@@ -458,6 +459,7 @@ class ArvadosContainer(JobBase):
     def done(self, record):
         outputs = {}
         retried = False
+        rcode = None
         try:
             container = self.arvrunner.api.containers().get(
                 uuid=record["container_uuid"]
@@ -476,7 +478,7 @@ class ArvadosContainer(JobBase):
                     processStatus = "permanentFail"
 
                 if processStatus == "permanentFail" and self.attempt_count == 1 and self.out_of_memory_retry(record, container):
-                    logger.info("%s Container failed with out of memory error, retrying with more RAM.",
+                    logger.warning("%s Container failed with out of memory error, retrying with more RAM.",
                                  self.arvrunner.label(self))
                     self.job_runtime.submit_request_uuid = None
                     self.uuid = None
@@ -485,7 +487,7 @@ class ArvadosContainer(JobBase):
                     return
 
                 if rcode == 137:
-                    logger.warning("%s Container may have been killed for using too much RAM.  Try resubmitting with a higher 'ramMin'.",
+                    logger.warning("%s Container may have been killed for using too much RAM.  Try resubmitting with a higher 'ramMin' or use the arv:OutOfMemoryRetry feature.",
                                  self.arvrunner.label(self))
             else:
                 processStatus = "permanentFail"
@@ -498,7 +500,7 @@ class ArvadosContainer(JobBase):
                 label = self.arvrunner.label(self)
                 done.logtail(
                     logc, logger.error,
-                    "%s (%s) error log:" % (label, record["uuid"]), maxlen=40)
+                    "%s (%s) error log:" % (label, record["uuid"]), maxlen=40, include_crunchrun=(rcode is None or rcode > 127))
 
             if record["output_uuid"]:
                 if self.arvrunner.trash_intermediate or self.arvrunner.intermediate_output_ttl:
@@ -609,6 +611,8 @@ class RunnerContainer(Runner):
                 "content": packed
             }
             container_req["properties"]["template_uuid"] = self.embedded_tool.tool["id"][6:33]
+        elif self.embedded_tool.tool.get("id", "").startswith("file:"):
+            raise Exception("Tool id '%s' is a local file but expected keep: or arvwf:" % self.embedded_tool.tool.get("id"))
         else:
             main = self.loadingContext.loader.idx["_:main"]
             if main.get("id") == "_:main":