def test_serialization_settings_transport():
default_img = Image(name="default", fqn="test", tag="tag")
serialization_settings = SerializationSettings(
project="project",
domain="domain",
version="version",
env={"hello": "blah"},
image_config=ImageConfig(
default_image=default_img,
images=[default_img],
),
flytekit_virtualenv_root="/opt/venv/blah",
python_interpreter="/opt/venv/bin/python3",
fast_serialization_settings=FastSerializationSettings(
enabled=True,
destination_dir="/opt/blah/blah/blah",
distribution_location="s3://my-special-bucket/blah/bha/asdasdasd/cbvsdsdf/asdddasdasdasdasdasdasd.tar.gz",
),
)
tp = serialization_settings.serialized_context
with_serialized = serialization_settings.with_serialized_context()
assert serialization_settings.env == {"hello": "blah"}
assert with_serialized.env
assert with_serialized.env[SERIALIZED_CONTEXT_ENV_VAR] == tp
ss = SerializationSettings.from_transport(tp)
assert ss is not None
assert ss == serialization_settings
assert len(tp) == 376
def setup_execution(
raw_output_data_prefix: str,
checkpoint_path: Optional[str] = None,
prev_checkpoint: Optional[str] = None,
dynamic_addl_distro: Optional[str] = None,
dynamic_dest_dir: Optional[str] = None,
):
"""
:param raw_output_data_prefix:
:param checkpoint_path:
:param prev_checkpoint:
:param dynamic_addl_distro: Works in concert with the other dynamic arg. If present, indicates that if a dynamic
task were to run, it should set fast serialize to true and use these values in FastSerializationSettings
:param dynamic_dest_dir: See above.
:return:
"""
exe_project = get_one_of("FLYTE_INTERNAL_EXECUTION_PROJECT", "_F_PRJ")
exe_domain = get_one_of("FLYTE_INTERNAL_EXECUTION_DOMAIN", "_F_DM")
exe_name = get_one_of("FLYTE_INTERNAL_EXECUTION_ID", "_F_NM")
exe_wf = get_one_of("FLYTE_INTERNAL_EXECUTION_WORKFLOW", "_F_WF")
exe_lp = get_one_of("FLYTE_INTERNAL_EXECUTION_LAUNCHPLAN", "_F_LP")
tk_project = get_one_of("FLYTE_INTERNAL_TASK_PROJECT", "_F_TK_PRJ")
tk_domain = get_one_of("FLYTE_INTERNAL_TASK_DOMAIN", "_F_TK_DM")
tk_name = get_one_of("FLYTE_INTERNAL_TASK_NAME", "_F_TK_NM")
tk_version = get_one_of("FLYTE_INTERNAL_TASK_VERSION", "_F_TK_V")
compressed_serialization_settings = os.environ.get(SERIALIZED_CONTEXT_ENV_VAR, "")
ctx = FlyteContextManager.current_context()
# Create directories
user_workspace_dir = ctx.file_access.get_random_local_directory()
logger.info(f"Using user directory {user_workspace_dir}")
pathlib.Path(user_workspace_dir).mkdir(parents=True, exist_ok=True)
from flytekit import __version__ as _api_version
checkpointer = None
if checkpoint_path is not None:
checkpointer = SyncCheckpoint(checkpoint_dest=checkpoint_path, checkpoint_src=prev_checkpoint)
logger.debug(f"Checkpointer created with source {prev_checkpoint} and dest {checkpoint_path}")
execution_parameters = ExecutionParameters(
execution_id=_identifier.WorkflowExecutionIdentifier(
project=exe_project,
domain=exe_domain,
name=exe_name,
),
execution_date=_datetime.datetime.utcnow(),
stats=_get_stats(
cfg=StatsConfig.auto(),
# Stats metric path will be:
# registration_project.registration_domain.app.module.task_name.user_stats
# and it will be tagged with execution-level values for project/domain/wf/lp
prefix=f"{tk_project}.{tk_domain}.{tk_name}.user_stats",
tags={
"exec_project": exe_project,
"exec_domain": exe_domain,
"exec_workflow": exe_wf,
"exec_launchplan": exe_lp,
"api_version": _api_version,
},
),
logging=user_space_logger,
tmp_dir=user_workspace_dir,
raw_output_prefix=raw_output_data_prefix,
checkpoint=checkpointer,
task_id=_identifier.Identifier(_identifier.ResourceType.TASK, tk_project, tk_domain, tk_name, tk_version),
)
try:
file_access = FileAccessProvider(
local_sandbox_dir=tempfile.mkdtemp(prefix="flyte"),
raw_output_prefix=raw_output_data_prefix,
)
except TypeError: # would be thrown from DataPersistencePlugins.find_plugin
logger.error(f"No data plugin found for raw output prefix {raw_output_data_prefix}")
raise
es = ctx.new_execution_state().with_params(
mode=ExecutionState.Mode.TASK_EXECUTION,
user_space_params=execution_parameters,
)
cb = ctx.new_builder().with_file_access(file_access).with_execution_state(es)
if compressed_serialization_settings:
ss = SerializationSettings.from_transport(compressed_serialization_settings)
ssb = ss.new_builder()
ssb.project = exe_project
ssb.domain = exe_domain
ssb.version = tk_version
if dynamic_addl_distro:
ssb.fast_serialization_settings = FastSerializationSettings(
enabled=True,
destination_dir=dynamic_dest_dir,
distribution_location=dynamic_addl_distro,
)
cb = cb.with_serialization_settings(ssb.build())
with FlyteContextManager.with_context(cb) as ctx:
yield ctx