X-Git-Url: https://git.arvados.org/arvados.git/blobdiff_plain/1d387fdfc68ba8c07ab42aa104ab3a8369e7d0ec..54050447a56ab3e019746f12eb36b026dd5d2fb0:/crunch_scripts/run-command diff --git a/crunch_scripts/run-command b/crunch_scripts/run-command index 7a65f53773..e6ec889bbc 100755 --- a/crunch_scripts/run-command +++ b/crunch_scripts/run-command @@ -49,7 +49,7 @@ subst.default_subs["node.cores"] = sub_cores rcode = 1 def machine_progress(bytes_written, bytes_expected): - return "run-command: {} written {} total\n".format( + return "run-command: wrote {} total {}\n".format( bytes_written, -1 if (bytes_expected is None) else bytes_expected) class SigHandler(object): @@ -86,50 +86,49 @@ try: if sig.sig != None: print("run-command: terminating on signal %s" % sig.sig) - sys.exit(rcode) + sys.exit(2) else: print("run-command: completed with exit code %i (%s)" % (rcode, "success" if rcode == 0 else "failed")) except Exception as e: print("run-command: caught exception: {}".format(e)) -finally: - # restore default signal handlers. - signal.signal(signal.SIGINT, signal.SIG_DFL) - signal.signal(signal.SIGTERM, signal.SIG_DFL) - signal.signal(signal.SIGQUIT, signal.SIG_DFL) - - for l in links: - os.unlink(l) - - print("run-command: the follow output files will be saved to keep:") - - subprocess.call(["find", ".", "-type", "f", "-printf", "run-command: %12.12s %h/%f\\n"]) - - print("run-command: start writing output to keep") - - done = False - resume_cache = put.ResumeCache(os.path.join(arvados.current_task().tmpdir, "upload-output-checkpoint")) - reporter = put.progress_writer(machine_progress) - bytes_expected = put.expected_bytes_for(".") - while not done: - try: - out = put.ArvPutCollectionWriter(resume_cache, reporter, bytes_expected) - out.do_queued_work() - out.write_directory_tree(".", max_manifest_depth=0) - outuuid = out.finish() - api.job_tasks().update(uuid=arvados.current_task()['uuid'], - body={ - 'output':outuuid, - 'success': (rcode == 0), - 'progress':1.0 - }).execute() - done = True - except KeyboardInterrupt: - print("run-command: terminating on signal SIGINT") - done = True - except Exception as e: - print("run-command: caught exception: {}".format(e)) - time.sleep(5) +# restore default signal handlers. +signal.signal(signal.SIGINT, signal.SIG_DFL) +signal.signal(signal.SIGTERM, signal.SIG_DFL) +signal.signal(signal.SIGQUIT, signal.SIG_DFL) + +for l in links: + os.unlink(l) + +print("run-command: the follow output files will be saved to keep:") + +subprocess.call(["find", ".", "-type", "f", "-printf", "run-command: %12.12s %h/%f\\n"]) + +print("run-command: start writing output to keep") + +done = False +resume_cache = put.ResumeCache(os.path.join(arvados.current_task().tmpdir, "upload-output-checkpoint")) +reporter = put.progress_writer(machine_progress) +bytes_expected = put.expected_bytes_for(".") +while not done: + try: + out = put.ArvPutCollectionWriter.from_cache(resume_cache, reporter, bytes_expected) + out.do_queued_work() + out.write_directory_tree(".", max_manifest_depth=0) + outuuid = out.finish() + api.job_tasks().update(uuid=arvados.current_task()['uuid'], + body={ + 'output':outuuid, + 'success': (rcode == 0), + 'progress':1.0 + }).execute() + done = True + except KeyboardInterrupt: + print("run-command: terminating on signal 2") + sys.exit(2) + except Exception as e: + print("run-command: caught exception: {}".format(e)) + time.sleep(5) sys.exit(rcode)