sdk/cwl/arvados_cwl/runner.py

   1 import os
   2 import urlparse
   3 from functools import partial
   4 import logging
   5 import json
   6 import re
   7 import subprocess
   8
   9 from StringIO import StringIO
  10
  11 from schema_salad.sourceline import SourceLine
  12
  13 import cwltool.draft2tool
  14 from cwltool.draft2tool import CommandLineTool
  15 import cwltool.workflow
  16 from cwltool.process import get_feature, scandeps, UnsupportedRequirement, normalizeFilesDirs, shortname
  17 from cwltool.load_tool import fetch_document
  18 from cwltool.pathmapper import adjustFileObjs, adjustDirObjs
  19 from cwltool.utils import aslist
  20 from cwltool.builder import substitute
  21 from cwltool.pack import pack
  22
  23 import arvados.collection
  24 import ruamel.yaml as yaml
  25
  26 from .arvdocker import arv_docker_get_image
  27 from .pathmapper import ArvPathMapper
  28 from ._version import __version__
  29 from . import done
  30
  31 logger = logging.getLogger('arvados.cwl-runner')
  32
  33 cwltool.draft2tool.ACCEPTLIST_RE = re.compile(r".*")
  34
  35 def trim_listing(obj):
  36     """Remove 'listing' field from Directory objects that are keep references.
  37
  38     When Directory objects represent Keep references, it redundant and
  39     potentially very expensive to pass fully enumerated Directory objects
  40     between instances of cwl-runner (e.g. a submitting a job, or using the
  41     RunInSingleContainer feature), so delete the 'listing' field when it is
  42     safe to do so.
  43     """
  44
  45     if obj.get("location", "").startswith("keep:") and "listing" in obj:
  46         del obj["listing"]
  47     if obj.get("location", "").startswith("_:"):
  48         del obj["location"]
  49
  50 def upload_dependencies(arvrunner, name, document_loader,
  51                         workflowobj, uri, loadref_run, include_primary=True):
  52     """Upload the dependencies of the workflowobj document to Keep.
  53
  54     Returns a pathmapper object mapping local paths to keep references.  Also
  55     does an in-place update of references in "workflowobj".
  56
  57     Use scandeps to find $import, $include, $schemas, run, File and Directory
  58     fields that represent external references.
  59
  60     If workflowobj has an "id" field, this will reload the document to ensure
  61     it is scanning the raw document prior to preprocessing.
  62     """
  63
  64     loaded = set()
  65     def loadref(b, u):
  66         joined = document_loader.fetcher.urljoin(b, u)
  67         defrg, _ = urlparse.urldefrag(joined)
  68         if defrg not in loaded:
  69             loaded.add(defrg)
  70             # Use fetch_text to get raw file (before preprocessing).
  71             text = document_loader.fetch_text(defrg)
  72             if isinstance(text, bytes):
  73                 textIO = StringIO(text.decode('utf-8'))
  74             else:
  75                 textIO = StringIO(text)
  76             return yaml.safe_load(textIO)
  77         else:
  78             return {}
  79
  80     if loadref_run:
  81         loadref_fields = set(("$import", "run"))
  82     else:
  83         loadref_fields = set(("$import",))
  84
  85     scanobj = workflowobj
  86     if "id" in workflowobj:
  87         # Need raw file content (before preprocessing) to ensure
  88         # that external references in $include and $mixin are captured.
  89         scanobj = loadref("", workflowobj["id"])
  90
  91     sc = scandeps(uri, scanobj,
  92                   loadref_fields,
  93                   set(("$include", "$schemas", "location")),
  94                   loadref, urljoin=document_loader.fetcher.urljoin)
  95
  96     normalizeFilesDirs(sc)
  97
  98     if include_primary and "id" in workflowobj:
  99         sc.append({"class": "File", "location": workflowobj["id"]})
 100
 101     if "$schemas" in workflowobj:
 102         for s in workflowobj["$schemas"]:
 103             sc.append({"class": "File", "location": s})
 104
 105     mapper = ArvPathMapper(arvrunner, sc, "",
 106                            "keep:%s",
 107                            "keep:%s/%s",
 108                            name=name)
 109
 110     def setloc(p):
 111         if "location" in p and (not p["location"].startswith("_:")) and (not p["location"].startswith("keep:")):
 112             p["location"] = mapper.mapper(p["location"]).resolved
 113     adjustFileObjs(workflowobj, setloc)
 114     adjustDirObjs(workflowobj, setloc)
 115
 116     if "$schemas" in workflowobj:
 117         sch = []
 118         for s in workflowobj["$schemas"]:
 119             sch.append(mapper.mapper(s).resolved)
 120         workflowobj["$schemas"] = sch
 121
 122     return mapper
 123
 124
 125 def upload_docker(arvrunner, tool):
 126     """Visitor which uploads Docker images referenced in CommandLineTool objects."""
 127     if isinstance(tool, CommandLineTool):
 128         (docker_req, docker_is_req) = get_feature(tool, "DockerRequirement")
 129         if docker_req:
 130             if docker_req.get("dockerOutputDirectory"):
 131                 # TODO: can be supported by containers API, but not jobs API.
 132                 raise SourceLine(docker_req, "dockerOutputDirectory", UnsupportedRequirement).makeError(
 133                     "Option 'dockerOutputDirectory' of DockerRequirement not supported.")
 134             arv_docker_get_image(arvrunner.api, docker_req, True, arvrunner.project_uuid)
 135
 136 def packed_workflow(arvrunner, tool):
 137     """Create a packed workflow.
 138
 139     A "packed" workflow is one where all the components have been combined into a single document."""
 140
 141     return pack(tool.doc_loader, tool.doc_loader.fetch(tool.tool["id"]),
 142                 tool.tool["id"], tool.metadata)
 143
 144 def tag_git_version(packed):
 145     if tool.tool["id"].startswith("file://"):
 146         path = os.path.dirname(tool.tool["id"][7:])
 147         try:
 148             githash = subprocess.check_output(['git', 'log', '--first-parent', '--max-count=1', '--format=%H'], stderr=subprocess.STDOUT, cwd=path).strip()
 149         except (OSError, subprocess.CalledProcessError):
 150             pass
 151         else:
 152             packed["http://schema.org/version"] = githash
 153
 154
 155 def upload_job_order(arvrunner, name, tool, job_order):
 156     """Upload local files referenced in the input object and return updated input
 157     object with 'location' updated to the proper keep references.
 158     """
 159
 160     for t in tool.tool["inputs"]:
 161         def setSecondary(fileobj):
 162             if isinstance(fileobj, dict) and fileobj.get("class") == "File":
 163                 if "secondaryFiles" not in fileobj:
 164                     fileobj["secondaryFiles"] = [{"location": substitute(fileobj["location"], sf), "class": "File"} for sf in t["secondaryFiles"]]
 165
 166             if isinstance(fileobj, list):
 167                 for e in fileobj:
 168                     setSecondary(e)
 169
 170         if shortname(t["id"]) in job_order and t.get("secondaryFiles"):
 171             setSecondary(job_order[shortname(t["id"])])
 172
 173     jobmapper = upload_dependencies(arvrunner,
 174                                     name,
 175                                     tool.doc_loader,
 176                                     job_order,
 177                                     job_order.get("id", "#"),
 178                                     False)
 179
 180     if "id" in job_order:
 181         del job_order["id"]
 182
 183     # Need to filter this out, gets added by cwltool when providing
 184     # parameters on the command line.
 185     if "job_order" in job_order:
 186         del job_order["job_order"]
 187
 188     return job_order
 189
 190 def upload_workflow_deps(arvrunner, tool):
 191     # Ensure that Docker images needed by this workflow are available
 192     tool.visit(partial(upload_docker, arvrunner))
 193
 194     document_loader = tool.doc_loader
 195
 196     def upload_tool_deps(deptool):
 197         if "id" in deptool:
 198             upload_dependencies(arvrunner,
 199                                 "%s dependencies" % (shortname(deptool["id"])),
 200                                 document_loader,
 201                                 deptool,
 202                                 deptool["id"],
 203                                 False,
 204                                 include_primary=False)
 205             document_loader.idx[deptool["id"]] = deptool
 206
 207     tool.visit(upload_tool_deps)
 208
 209 def arvados_jobs_image(arvrunner, img):
 210     """Determine if the right arvados/jobs image version is available.  If not, try to pull and upload it."""
 211
 212     try:
 213         arv_docker_get_image(arvrunner.api, {"dockerPull": img}, True, arvrunner.project_uuid)
 214     except Exception as e:
 215         raise Exception("Docker image %s is not available\n%s" % (img, e) )
 216     return img
 217
 218 def upload_workflow_collection(arvrunner, name, packed):
 219     collection = arvados.collection.Collection(api_client=arvrunner.api,
 220                                                keep_client=arvrunner.keep_client,
 221                                                num_retries=arvrunner.num_retries)
 222     with collection.open("workflow.cwl", "w") as f:
 223         f.write(json.dumps(packed, indent=2, sort_keys=True, separators=(',',': ')))
 224
 225     filters = [["portable_data_hash", "=", collection.portable_data_hash()],
 226                ["name", "like", name+"%"]]
 227     if arvrunner.project_uuid:
 228         filters.append(["owner_uuid", "=", arvrunner.project_uuid])
 229     exists = arvrunner.api.collections().list(filters=filters).execute(num_retries=arvrunner.num_retries)
 230
 231     if exists["items"]:
 232         logger.info("Using collection %s", exists["items"][0]["uuid"])
 233     else:
 234         collection.save_new(name=name,
 235                             owner_uuid=arvrunner.project_uuid,
 236                             ensure_unique_name=True,
 237                             num_retries=arvrunner.num_retries)
 238         logger.info("Uploaded to %s", collection.manifest_locator())
 239
 240     return collection.portable_data_hash()
 241
 242
 243 class Runner(object):
 244     """Base class for runner processes, which submit an instance of
 245     arvados-cwl-runner and wait for the final result."""
 246
 247     def __init__(self, runner, tool, job_order, enable_reuse,
 248                  output_name, output_tags, submit_runner_ram=0,
 249                  name=None, on_error=None, submit_runner_image=None):
 250         self.arvrunner = runner
 251         self.tool = tool
 252         self.job_order = job_order
 253         self.running = False
 254         self.enable_reuse = enable_reuse
 255         self.uuid = None
 256         self.final_output = None
 257         self.output_name = output_name
 258         self.output_tags = output_tags
 259         self.name = name
 260         self.on_error = on_error
 261         self.jobs_image = submit_runner_image or "arvados/jobs:"+__version__
 262
 263         if submit_runner_ram:
 264             self.submit_runner_ram = submit_runner_ram
 265         else:
 266             self.submit_runner_ram = 1024
 267
 268         if self.submit_runner_ram <= 0:
 269             raise Exception("Value of --submit-runner-ram must be greater than zero")
 270
 271     def update_pipeline_component(self, record):
 272         pass
 273
 274     def done(self, record):
 275         """Base method for handling a completed runner."""
 276
 277         try:
 278             if record["state"] == "Complete":
 279                 if record.get("exit_code") is not None:
 280                     if record["exit_code"] == 33:
 281                         processStatus = "UnsupportedRequirement"
 282                     elif record["exit_code"] == 0:
 283                         processStatus = "success"
 284                     else:
 285                         processStatus = "permanentFail"
 286                 else:
 287                     processStatus = "success"
 288             else:
 289                 processStatus = "permanentFail"
 290
 291             outputs = {}
 292
 293             if processStatus == "permanentFail":
 294                 logc = arvados.collection.CollectionReader(record["log"],
 295                                                            api_client=self.arvrunner.api,
 296                                                            keep_client=self.arvrunner.keep_client,
 297                                                            num_retries=self.arvrunner.num_retries)
 298                 done.logtail(logc, logger, "%s error log:" % self.arvrunner.label(self), maxlen=40)
 299
 300             self.final_output = record["output"]
 301             outc = arvados.collection.CollectionReader(self.final_output,
 302                                                        api_client=self.arvrunner.api,
 303                                                        keep_client=self.arvrunner.keep_client,
 304                                                        num_retries=self.arvrunner.num_retries)
 305             if "cwl.output.json" in outc:
 306                 with outc.open("cwl.output.json") as f:
 307                     if f.size() > 0:
 308                         outputs = json.load(f)
 309             def keepify(fileobj):
 310                 path = fileobj["location"]
 311                 if not path.startswith("keep:"):
 312                     fileobj["location"] = "keep:%s/%s" % (record["output"], path)
 313             adjustFileObjs(outputs, keepify)
 314             adjustDirObjs(outputs, keepify)
 315         except Exception as e:
 316             logger.exception("[%s] While getting final output object: %s", self.name, e)
 317             self.arvrunner.output_callback({}, "permanentFail")
 318         else:
 319             self.arvrunner.output_callback(outputs, processStatus)
 320         finally:
 321             if record["uuid"] in self.arvrunner.processes:
 322                 del self.arvrunner.processes[record["uuid"]]