Stream SQLite databases directly to Borg instead of dumping to an intermediate file (#807).

NEWS

@@ -6,6 +6,7 @@
  * #800: Add configured repository labels to the JSON output for all actions.
  * #802: The "check --force" flag now runs checks even if "check" is in "skip_actions".
  * #804: Validate the configured action names in the "skip_actions" option.
+ * #807: Stream SQLite databases directly to Borg instead of dumping to an intermediate file.
  * When logging commands that borgmatic executes, log the environment variables that
    borgmatic sets for those commands. (But don't log their values, since they often contain
    passwords.)

borgmatic/hooks/sqlite.py

@@ -18,10 +18,12 @@ def make_dump_path(config):  # pragma: no cover
 
 def dump_data_sources(databases, config, log_prefix, dry_run):
     '''
-    Dump the given SQLite3 databases to a file. The databases are supplied as a sequence of
+    Dump the given SQLite3 databases to a named pipe. The databases are supplied as a sequence of
     configuration dicts, as per the configuration schema. Use the given configuration dict to
-    construct the destination path and the given log prefix in any log entries. If this is a dry
-    run, then don't actually dump anything.
+    construct the destination path and the given log prefix in any log entries.
+
+    Return a sequence of subprocess.Popen instances for the dump processes ready to spew to a named
+    pipe. But if this is a dry run, then don't actually dump anything and return an empty sequence.
     '''
     dry_run_label = ' (dry run; not actually dumping anything)' if dry_run else ''
     processes = []
@@ -40,6 +42,7 @@ def dump_data_sources(databases, config, log_prefix, dry_run):
 
         dump_path = make_dump_path(config)
         dump_filename = dump.make_data_source_dump_filename(dump_path, database['name'])
+
         if os.path.exists(dump_filename):
             logger.warning(
                 f'{log_prefix}: Skipping duplicate dump of SQLite database at {database_path} to {dump_filename}'
@@ -59,7 +62,7 @@ def dump_data_sources(databases, config, log_prefix, dry_run):
         if dry_run:
             continue
 
-        dump.create_parent_directory_for_dump(dump_filename)
+        dump.create_named_pipe_for_dump(dump_filename)
         processes.append(execute_command(command, shell=True, run_to_completion=False))
 
     return processes

tests/unit/hooks/test_sqlite.py

@@ -13,7 +13,7 @@ def test_dump_data_sources_logs_and_skips_if_dump_already_exists():
         '/path/to/dump/database'
     )
     flexmock(module.os.path).should_receive('exists').and_return(True)
-    flexmock(module.dump).should_receive('create_parent_directory_for_dump').never()
+    flexmock(module.dump).should_receive('create_named_pipe_for_dump').never()
     flexmock(module).should_receive('execute_command').never()
 
     assert module.dump_data_sources(databases, {}, 'test.yaml', dry_run=False) == []
@@ -31,7 +31,7 @@ def test_dump_data_sources_dumps_each_database():
         '/path/to/dump/database'
     )
     flexmock(module.os.path).should_receive('exists').and_return(False)
-    flexmock(module.dump).should_receive('create_parent_directory_for_dump')
+    flexmock(module.dump).should_receive('create_named_pipe_for_dump')
     flexmock(module).should_receive('execute_command').and_return(processes[0]).and_return(
         processes[1]
     )
@@ -39,7 +39,7 @@ def test_dump_data_sources_dumps_each_database():
     assert module.dump_data_sources(databases, {}, 'test.yaml', dry_run=False) == processes
 
 
-def test_dumping_database_with_non_existent_path_warns_and_dumps_database():
+def test_dump_data_sources_with_non_existent_path_warns_and_dumps_database():
     databases = [
         {'path': '/path/to/database1', 'name': 'database1'},
     ]
@@ -51,13 +51,13 @@ def test_dumping_database_with_non_existent_path_warns_and_dumps_database():
         '/path/to/dump/database'
     )
     flexmock(module.os.path).should_receive('exists').and_return(False)
-    flexmock(module.dump).should_receive('create_parent_directory_for_dump')
+    flexmock(module.dump).should_receive('create_named_pipe_for_dump')
     flexmock(module).should_receive('execute_command').and_return(processes[0])
 
     assert module.dump_data_sources(databases, {}, 'test.yaml', dry_run=False) == processes
 
 
-def test_dumping_database_with_name_all_warns_and_dumps_all_databases():
+def test_dump_data_sources_with_name_all_warns_and_dumps_all_databases():
     databases = [
         {'path': '/path/to/database1', 'name': 'all'},
     ]
@@ -71,7 +71,7 @@ def test_dumping_database_with_name_all_warns_and_dumps_all_databases():
         '/path/to/dump/database'
     )
     flexmock(module.os.path).should_receive('exists').and_return(False)
-    flexmock(module.dump).should_receive('create_parent_directory_for_dump')
+    flexmock(module.dump).should_receive('create_named_pipe_for_dump')
     flexmock(module).should_receive('execute_command').and_return(processes[0])
 
     assert module.dump_data_sources(databases, {}, 'test.yaml', dry_run=False) == processes
@@ -85,7 +85,7 @@ def test_dump_data_sources_does_not_dump_if_dry_run():
         '/path/to/dump/database'
     )
     flexmock(module.os.path).should_receive('exists').and_return(False)
-    flexmock(module.dump).should_receive('create_parent_directory_for_dump').never()
+    flexmock(module.dump).should_receive('create_named_pipe_for_dump').never()
     flexmock(module).should_receive('execute_command').never()
 
     assert module.dump_data_sources(databases, {}, 'test.yaml', dry_run=True) == []