PsqlDos: Add migration to remove hashes for all CalcJobNodes

The algorithm for computing the hashes of `CalcJobNodes` contained a bug
which was fixed in 685e0f8. This means
however that all existing hashes for `CalcJobNodes` are inconsistent
with the new algorithm. This would mean that valid cache hits would be
missed, and worse, different calculatons could end up with the same hash
and mistakingly be cached from one another.

The migration drops the `_aiida_hash` extra which is used to store the
node's hash. A warning is emitted to notify the user, suggesting to run
`verdi node rehash` to recompute the hash of all `CalcJobNodes`.

aiida/storage/psql_dos/migrations/versions/main_0002_recompute_hash_calc_job_node.py

@@ -0,0 +1,42 @@
+# -*- coding: utf-8 -*-
+###########################################################################
+# Copyright (c), The AiiDA team. All rights reserved.                     #
+# This file is part of the AiiDA code.                                    #
+#                                                                         #
+# The code is hosted on GitHub at https://github.com/aiidateam/aiida-core #
+# For further information on the license, see the LICENSE.txt file        #
+# For further information please visit http://www.aiida.net               #
+###########################################################################
+# pylint: disable=invalid-name,no-member
+"""Drop the hashes for all ``CalcJobNode`` instances.
+
+The computed hash erroneously included the hash of the file repository. This was present as of v2.0 and so all nodes
+created with versions since then will have incorrect hashes.
+
+Revision ID: main_0002
+Revises: main_0001
+Create Date: 2023-05-05
+
+"""
+from alembic import op
+
+from aiida.storage.psql_dos.migrations.utils.integrity import drop_hashes
+
+revision = 'main_0002'
+down_revision = 'main_0001'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    """Migrations for the upgrade."""
+    drop_hashes(
+        op.get_bind(), hash_extra_key='_aiida_hash', entry_point_string='aiida.node:process.calculation.calcjob'
+    )
+
+
+def downgrade():
+    """Migrations for the downgrade."""
+    drop_hashes(
+        op.get_bind(), hash_extra_key='_aiida_hash', entry_point_string='aiida.node:process.calculation.calcjob'
+    )

tests/storage/psql_dos/migrations/main_branch/test_main_0002_recompute_hash_calc_job_node.py

@@ -0,0 +1,76 @@
+# -*- coding: utf-8 -*-
+###########################################################################
+# Copyright (c), The AiiDA team. All rights reserved.                     #
+# This file is part of the AiiDA code.                                    #
+#                                                                         #
+# The code is hosted on GitHub at https://github.com/aiidateam/aiida-core #
+# For further information on the license, see the LICENSE.txt file        #
+# For further information please visit http://www.aiida.net               #
+###########################################################################
+"""Test ``main_0002_recompute_hash_calc_job_node.py``."""
+from aiida.common import timezone
+from aiida.common.utils import get_new_uuid
+from aiida.storage.psql_dos.migrator import PsqlDosMigrator
+
+
+def test_migration(perform_migrations: PsqlDosMigrator):
+    """Test the migration removes the ``_aiida_hash`` extra of all ``CalcJobNodes`` and those nodes only."""
+    perform_migrations.migrate_up('main@main_0001')
+
+    user_model = perform_migrations.get_current_table('db_dbuser')
+    node_model = perform_migrations.get_current_table('db_dbnode')
+
+    with perform_migrations.session() as session:
+        user = user_model(email='test', first_name='test', last_name='test', institution='test')
+        session.add(user)
+        session.commit()
+
+        calcjob = node_model(
+            uuid=get_new_uuid(),
+            user_id=user.id,
+            ctime=timezone.now(),
+            mtime=timezone.now(),
+            label='test',
+            description='',
+            node_type='process.calculation.calcjob.CalcJobNode.',
+            process_type='aiida.calculations:core.arithmetic.add',
+            attributes={'parser_name': 'core.arithmetic.add'},
+            repository_metadata={},
+            extras={
+                '_aiida_hash': 'hash',
+                'other_extra': 'value'
+            }
+        )
+        workflow = node_model(
+            uuid=get_new_uuid(),
+            user_id=user.id,
+            ctime=timezone.now(),
+            mtime=timezone.now(),
+            label='test',
+            description='',
+            node_type='process.workflow.workchain.WorkChainNode.',
+            process_type='aiida.workflows:core.arithmetic.add_multiply',
+            repository_metadata={},
+            extras={
+                '_aiida_hash': 'hash',
+                'other_extra': 'value'
+            }
+        )
+        session.add_all((calcjob, workflow))
+        session.commit()
+
+        calcjob_id = calcjob.id
+        workflow_id = workflow.id
+
+    # Perform the migration that is being tested.
+    perform_migrations.migrate_up('main@main_0002')
+
+    node_model = perform_migrations.get_current_table('db_dbnode')
+
+    # Check that the ``_aiida_hash`` extra was removed, and only that extra, and only for the calcjob, not the workflow.
+    with perform_migrations.session() as session:
+        calcjob = session.query(node_model).filter(node_model.id == calcjob_id).one()
+        assert calcjob.extras == {'other_extra': 'value'}
+
+        workflow = session.query(node_model).filter(node_model.id == workflow_id).one()
+        assert workflow.extras == {'_aiida_hash': 'hash', 'other_extra': 'value'}