diff --git a/extras/import_old_redings.py b/extras/import_old_redings.py
new file mode 100644
index 0000000..f1f3e97
--- /dev/null
+++ b/extras/import_old_redings.py
@@ -0,0 +1,54 @@
+
+import django
+from time import time
+from pyscada.models import Variable
+from concurrent.futures import ThreadPoolExecutor, ProcessPoolExecutor
+
+
+
+def subprocess_setup():
+    django.setup()
+    from django.db import connections
+    for conn in connections.all():
+        conn.close()
+
+def process_variables(variable_ids):
+    for variable_id in variable_ids:
+        v = Variable.objects.get(pk=variable_id)
+        points = []
+        i = 0
+        #          now     days                primary key offset, -> ms
+        min_time = (time()-(1 * 24 * 60 * 60)) * 2097152 * 1000
+        data = v.recordeddata_set.filter(pk__gt=min_time)
+        i_max = data.count()
+        if i_max == 0:
+            continue
+        print(f"{v} {i} {i_max} {i/i_max*100}")
+        if not hasattr(v.datasource,"influxdatabase"):
+            continue
+        new_point = v.datasource.influxdatabase.create_data_element_from_variable
+        for rc in data.iterator():
+            points.append(new_point(v ,rc.timestamp, rc.value()))
+            i += 1
+            if i%5000==0:
+                v.datasource.influxdatabase.write_points(points)
+                print(f"{v} {i} {i_max} {i/i_max*100:1.2f}")
+                points = []
+        v.datasource.influxdatabase.write_points(points)
+        print(f"{v} {i} {i_max} {i/i_max*100:1.2f}")
+        points = []
+
+variables = []
+workers = 1
+for worker_id in range(workers):
+    variables.append([])
+worker = 0
+
+for em in Variable.objects.all():
+    variables[worker % workers].append(em.pk)
+    worker += 1
+
+#with ProcessPoolExecutor(max_workers=workers, initializer=subprocess_setup) as executor:
+#    for worker_id in range(workers):
+#        executor.submit(process_variables, variables[worker_id])
+process_variables(variables[0])