Require date to reset DB

doc/source/user/projectconf.rst

@@ -239,28 +239,29 @@ Project specs
 
 .. _projectconf general:
 
-General Settings
-================
+General Settings - Environment variables
+========================================
 
 Aside from the project specific configuration, a few options can also be
 defined in general. There are two ways to set these options:
 
 * set the value in the ``~/.jfremote.yaml`` configuration file.
-* export the variable name prepended by the ``jfremote`` prefix::
+* set an environment variable composed by the name of the variable and
+  prepended by the ``JFREMOTE_`` prefix::
 
-    export jfremote_project=project_name
+    export JFREMOTE_PROJECT=project_name
 
 .. note::
 
-    The name of the exported variables is case-insensitive (i.e. JFREMOTE_PROJECT
+    The name of the exported variables is case-insensitive (i.e. jfremote_project
     is equally valid).
 
 The most useful variable to set is the ``project`` one, allowing to select the
 default project to be used in a multi-project environment.
 
 Other generic options are the location of the projects folder, instead of
-``~/.jfremote`` (``projects_folder``) and the path to the ``~/.jfremote.yaml``
-file itself (``config_file``).
+``~/.jfremote`` (``JFREMOTE_PROJECT_FOLDER``) and the path to the ``~/.jfremote.yaml``
+file itself (``JFREMOTE_CONFIG_FILE``).
 
 Some customization options are also available for the behaviour of the CLI.
 For more details see the API documentation :py:class:`jobflow_remote.config.settings.JobflowRemoteSettings`.

src/jobflow_remote/cli/admin.py

@@ -39,6 +39,16 @@
 
 @app_admin.command()
 def reset(
+    validation: Annotated[
+        Optional[str],
+        typer.Argument(
+            help=(
+                "If the number of flows in the DB exceed 25 it will be required to pass "
+                "today's date in the YYYY-MM-DD to proceed with the reset"
+            ),
+            metavar="DATE",
+        ),
+    ] = None,
     reset_output: Annotated[
         bool,
         typer.Option(
@@ -47,16 +57,6 @@ def reset(
             help="Also delete all the documents in the current store",
         ),
     ] = False,
-    max_limit: Annotated[
-        int,
-        typer.Option(
-            "--max",
-            "-m",
-            help=(
-                "The database will be reset only if the number of Flows is lower than the specified limit. 0 means no limit"
-            ),
-        ),
-    ] = 25,
     force: force_opt = False,
 ) -> None:
     """
@@ -81,45 +81,16 @@ def reset(
     with loading_spinner(processing=False) as progress:
         progress.add_task(description="Resetting the DB...", total=None)
         jc = get_job_controller()
-        done = jc.reset(reset_output=reset_output, max_limit=max_limit)
+        done = jc.reset(reset_output=reset_output, max_limit=25, validation=validation)
     not_text = "" if done else "[bold]NOT [/bold]"
     out_console.print(f"The database was {not_text}reset")
     if not done and SETTINGS.cli_suggestions:
         out_console.print(
-            "Check the amount of Flows and change --max-limit if this is the correct project to reset",
+            "Check the amount of Flows and set the DATE argument if this is the correct project to reset",
             style="yellow",
         )
 
 
-@app_admin.command(hidden=True)
-def remove_lock(
-    job_id: job_ids_indexes_opt = None,
-    db_id: db_ids_opt = None,
-    state: job_state_opt = None,
-    start_date: start_date_opt = None,
-    end_date: end_date_opt = None,
-    force: force_opt = False,
-) -> None:
-    """
-    DEPRECATED: use unlock instead
-    Forcibly removes the lock from the documents of the selected jobs.
-    WARNING: can lead to inconsistencies if the processes is actually running.
-    """
-    out_console.print(
-        "remove-lock command has been DEPRECATED. Use unlock instead.",
-        style="bold yellow",
-    )
-
-    unlock(
-        job_id=job_id,
-        db_id=db_id,
-        state=state,
-        start_date=start_date,
-        end_date=end_date,
-        force=force,
-    )
-
-
 @app_admin.command()
 def unlock(
     job_id: job_ids_indexes_opt = None,

src/jobflow_remote/jobs/jobcontroller.py

@@ -2516,7 +2516,12 @@ def unlock_flows(
         )
         return result.modified_count
 
-    def reset(self, reset_output: bool = False, max_limit: int = 25) -> bool:
+    def reset(
+        self,
+        reset_output: bool = False,
+        max_limit: int = 25,
+        validation: str | None = None,
+    ) -> bool:
         """
         Reset the content of the queue database and builds the indexes.
         Optionally deletes the content of the JobStore with the outputs.
@@ -2529,8 +2534,13 @@ def reset(self, reset_output: bool = False, max_limit: int = 25) -> bool:
             If True also reset the JobStore containing the outputs.
         max_limit
             Maximum number of Flows present in the DB. If number is larger
-            the database will not be reset. Set 0 for not limit.
-
+            the database a validation should be passed. Set 0 for not limit.
+            Setting max_limit to a large number or 0 will always lead to a
+            reset of the DB without validation. Prefer setting the password
+            to avoid unwanted deletions.
+        validation
+            A string representing today's date in the format YYYY-MM-DD.
+            Required if the number of Flows to delete exceed max_limit.
         Returns
         -------
         bool
@@ -2540,10 +2550,12 @@ def reset(self, reset_output: bool = False, max_limit: int = 25) -> bool:
         # what if the outputs are in other stores? Should take those as well
         if max_limit:
             n_flows = self.flows.count_documents({})
-            if n_flows >= max_limit:
+            today = datetime.now().strftime("%Y-%m-%d")
+            if n_flows >= max_limit and today != validation:
                 logger.warning(
                     f"The database contains {n_flows} flows and will not be reset. "
-                    "Increase the max_limit value or set it to 0"
+                    "Pass today's date in the YYYY-MM-DD format to validate the reset "
+                    "or change the max_limit value."
                 )
                 return False
 

tests/db/cli/test_admin.py

@@ -1,21 +1,41 @@
 import pytest
 
 
-def test_reset(job_controller, two_flows_four_jobs) -> None:
+def test_reset(job_controller, one_job) -> None:
+    from datetime import datetime
+
+    from jobflow import Flow
+
+    from jobflow_remote import submit_flow
+    from jobflow_remote.testing import add
     from jobflow_remote.testing.cli import run_check_cli
 
     run_check_cli(
-        ["admin", "reset", "-m", "1"],
+        ["admin", "reset"],
+        required_out="The database was reset",
+        cli_input="y",
+    )
+    assert job_controller.count_jobs() == 0
+
+    for _ in range(26):
+        f = Flow(add(1, 2))
+        submit_flow(f, worker="test_local_worker")
+
+    run_check_cli(
+        ["admin", "reset"],
         required_out="The database was NOT reset",
         cli_input="y",
     )
-    assert job_controller.count_jobs() == 4
+    assert job_controller.count_jobs() == 26
+
+    run_check_cli(["admin", "reset"], cli_input="n")
+    assert job_controller.count_jobs() == 26
 
-    run_check_cli(["admin", "reset", "-m", "10"], cli_input="n")
-    assert job_controller.count_jobs() == 4
+    run_check_cli(["admin", "reset", "1220-01-01"], cli_input="y")
+    assert job_controller.count_jobs() == 26
 
     run_check_cli(
-        ["admin", "reset", "-m", "10"],
+        ["admin", "reset", datetime.now().strftime("%Y-%m-%d")],
         required_out="The database was reset",
         cli_input="y",
     )

tests/db/conftest.py

@@ -164,7 +164,7 @@ def job_controller(random_project_name):
     from jobflow_remote.jobs.jobcontroller import JobController
 
     jc = JobController.from_project_name(random_project_name)
-    assert jc.reset()
+    assert jc.reset(max_limit=0)
     return jc
 
 

tests/db/jobs/test_jobcontroller.py

@@ -558,13 +558,33 @@ def test_set_job_doc_properties(job_controller, one_job) -> None:
 
 
 def test_reset(job_controller, two_flows_four_jobs):
+    from datetime import datetime
+
+    from jobflow import Flow
+
+    from jobflow_remote import submit_flow
+    from jobflow_remote.testing import add
+
     assert job_controller.count_jobs() == 4
 
     assert not job_controller.reset(max_limit=1)
+    assert job_controller.count_jobs() == 4
     assert job_controller.reset(max_limit=10, reset_output=True)
 
     assert job_controller.count_jobs() == 0
 
+    for _ in range(4):
+        f = Flow(add(1, 2))
+        submit_flow(f, worker="test_local_worker")
+
+    assert job_controller.count_jobs() == 4
+    assert not job_controller.reset(max_limit=1, validation="1327-01-01")
+    assert job_controller.count_jobs() == 4
+    assert job_controller.reset(
+        max_limit=1, validation=datetime.now().strftime("%Y-%m-%d")
+    )
+    assert job_controller.count_jobs() == 0
+
 
 def test_delete_job(job_controller, two_flows_four_jobs, runner):
     from jobflow import Flow

tests/integration/conftest.py

@@ -286,5 +286,5 @@ def job_controller(random_project_name):
     from jobflow_remote.jobs.jobcontroller import JobController
 
     jc = JobController.from_project_name(random_project_name)
-    assert jc.reset()
+    assert jc.reset(max_limit=0)
     return jc