studio: Move post_live_metrics from queue to executor (#8955)

* studio: Move `post_live_metrics` from `queue/workspace` to `executor/base`.

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Update dvc/repo/experiments/utils.py

Co-authored-by: skshetry <[email protected]>

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: skshetry <[email protected]>
This commit is contained in:
David de la Iglesia Castro
2023-02-06 13:31:10 +00:00
committed by GitHub
co-authored by skshetry pre-commit-ci[bot]
parent 80c2815e67
commit d84196db62
6 changed files with 65 additions and 54 deletions
+18
View File
@@ -32,6 +32,7 @@ from dvc.repo.experiments.refs import (
EXEC_CHECKPOINT,
ExpRefInfo,
)
from dvc.repo.experiments.utils import to_studio_params
from dvc.stage.serialize import to_lockfile
from dvc.ui import ui
from dvc.utils import dict_sha256, env2bool, relpath
@@ -545,6 +546,8 @@ class BaseExecutor(ABC):
log_errors: bool = True,
**kwargs,
):
from dvc_studio_client.post_live_metrics import post_live_metrics
from dvc.repo import Repo
from dvc.stage.monitor import CheckpointKilledError
@@ -561,6 +564,13 @@ class BaseExecutor(ABC):
os.chdir(dvc.root_dir)
try:
post_live_metrics(
"start",
info.baseline_rev,
info.name,
"dvc",
params=to_studio_params(dvc.params.show()),
)
logger.debug("Running repro in '%s'", os.getcwd())
yield dvc
info.status = TaskStatus.SUCCESS
@@ -578,6 +588,14 @@ class BaseExecutor(ABC):
info.status = TaskStatus.FAILED
raise
finally:
post_live_metrics(
"done",
info.baseline_rev,
info.name,
"dvc",
experiment_rev=dvc.experiments.scm.get_ref(EXEC_BRANCH),
)
if infofile is not None:
info.dump_json(infofile)
dvc.close()
+1 -17
View File
@@ -13,7 +13,7 @@ from dvc.repo.experiments.exceptions import ExpQueueEmptyError
from dvc.repo.experiments.executor.base import ExecutorInfo, TaskStatus
from dvc.repo.experiments.executor.local import WorkspaceExecutor
from dvc.repo.experiments.refs import EXEC_BRANCH, WORKSPACE_STASH
from dvc.repo.experiments.utils import get_exp_rwlock, to_studio_params
from dvc.repo.experiments.utils import get_exp_rwlock
from dvc.rwlock import RWLOCK_FILE, RWLOCK_LOCK, SCHEMA
from dvc.utils import relpath
from dvc.utils.fs import remove
@@ -94,8 +94,6 @@ class WorkspaceQueue(BaseStashQueue):
def _reproduce_entry(
self, entry: QueueEntry, executor: "BaseExecutor"
) -> Dict[str, Dict[str, str]]:
from dvc_studio_client.post_live_metrics import post_live_metrics
from dvc.stage.monitor import CheckpointKilledError
results: Dict[str, Dict[str, str]] = defaultdict(dict)
@@ -103,13 +101,6 @@ class WorkspaceQueue(BaseStashQueue):
infofile = self.get_infofile_path(exec_name)
try:
rev = entry.stash_rev
post_live_metrics(
"start",
executor.info.baseline_rev,
executor.info.name,
"dvc",
params=to_studio_params(self.repo.params.show()),
)
exec_result = executor.reproduce(
info=executor.info,
rev=rev,
@@ -131,13 +122,6 @@ class WorkspaceQueue(BaseStashQueue):
except Exception as exc: # noqa: BLE001
raise DvcException(f"Failed to reproduce experiment '{rev[:7]}'") from exc
finally:
post_live_metrics(
"done",
executor.info.baseline_rev,
executor.info.name,
"dvc",
experiment_rev=first(results.get(rev, {})),
)
executor.cleanup(infofile)
return results
+3 -1
View File
@@ -327,7 +327,9 @@ def to_studio_params(dvc_params):
"params.yaml": {"foo": 1}
}
"""
result = {}
result: Dict = {}
if not dvc_params:
return result
for rev_data in dvc_params.values():
for file_name, file_data in rev_data["data"].items():
result[file_name] = file_data["data"]
-36
View File
@@ -1,4 +1,3 @@
from dvc_studio_client import env
from funcy import first
@@ -28,38 +27,3 @@ def test_generate_random_exp_name(tmp_dir, dvc, scm, exp_stage, mocker):
# Can use same name because of different baseline_rev
ref = first(dvc.experiments.run(exp_stage.addressing, params=["foo=1"]))
assert dvc.experiments.get_exact_name([ref])[ref] == "0-0"
def test_post_to_studio(tmp_dir, dvc, scm, exp_stage, mocker, monkeypatch):
valid_response = mocker.MagicMock()
valid_response.status_code = 200
mocked_post = mocker.patch("requests.post", return_value=valid_response)
monkeypatch.setenv(env.STUDIO_ENDPOINT, "https://0.0.0.0")
monkeypatch.setenv(env.STUDIO_REPO_URL, "STUDIO_REPO_URL")
monkeypatch.setenv(env.STUDIO_TOKEN, "STUDIO_TOKEN")
baseline_sha = scm.get_rev()
exp_rev = first(dvc.experiments.run(exp_stage.addressing, params=["foo=1"]))
name = dvc.experiments.get_exact_name([exp_rev])[exp_rev]
assert mocked_post.call_count == 2
start_call, done_call = mocked_post.call_args_list
assert start_call.kwargs["json"] == {
"type": "start",
"repo_url": "STUDIO_REPO_URL",
"baseline_sha": baseline_sha,
"name": name,
"params": {"params.yaml": {"foo": 1}},
"client": "dvc",
}
assert done_call.kwargs["json"] == {
"type": "done",
"repo_url": "STUDIO_REPO_URL",
"baseline_sha": baseline_sha,
"name": name,
"client": "dvc",
"experiment_rev": exp_rev,
}
+2
View File
@@ -0,0 +1,2 @@
# pylint disable=unused-argument
from tests.unit.repo.experiments.conftest import exp_stage, test_queue # noqa: F401
@@ -0,0 +1,41 @@
import pytest
from dvc_studio_client import env
from funcy import first
@pytest.mark.parametrize("tmp", [True, False])
def test_post_to_studio(tmp_dir, dvc, scm, exp_stage, mocker, monkeypatch, tmp):
valid_response = mocker.MagicMock()
valid_response.status_code = 200
mocked_post = mocker.patch("requests.post", return_value=valid_response)
monkeypatch.setenv(env.STUDIO_ENDPOINT, "https://0.0.0.0")
monkeypatch.setenv(env.STUDIO_REPO_URL, "STUDIO_REPO_URL")
monkeypatch.setenv(env.STUDIO_TOKEN, "STUDIO_TOKEN")
baseline_sha = scm.get_rev()
exp_rev = first(
dvc.experiments.run(exp_stage.addressing, params=["foo=1"], tmp_dir=tmp)
)
name = dvc.experiments.get_exact_name([exp_rev])[exp_rev]
assert mocked_post.call_count == 2
start_call, done_call = mocked_post.call_args_list
assert start_call.kwargs["json"] == {
"type": "start",
"repo_url": "STUDIO_REPO_URL",
"baseline_sha": baseline_sha,
"name": name,
"params": {"params.yaml": {"foo": 1}},
"client": "dvc",
}
assert done_call.kwargs["json"] == {
"type": "done",
"repo_url": "STUDIO_REPO_URL",
"baseline_sha": baseline_sha,
"name": name,
"client": "dvc",
"experiment_rev": exp_rev,
}