diff --git a/xlib/sjob/__init__.py b/xlib/sjob/__init__.py
new file mode 100644
index 0000000..61771c9
--- /dev/null
+++ b/xlib/sjob/__init__.py
@@ -0,0 +1,9 @@
+"""
+Job processing in subprocesses.
+"""
+
+import multiprocessing
+if multiprocessing.get_start_method() != 'spawn':
+    multiprocessing.set_start_method("spawn", force=True)
+
+from .run_sequence import run_sequence
\ No newline at end of file
diff --git a/xlib/sjob/run_sequence.py b/xlib/sjob/run_sequence.py
new file mode 100644
index 0000000..02e04e0
--- /dev/null
+++ b/xlib/sjob/run_sequence.py
@@ -0,0 +1,91 @@
+import multiprocessing
+from typing import Callable, List
+
+from .. import console as lib_con
+
+
+def _run_sequence(barrier, init_func, init_kwargs, final_func, process_func, pipe):
+    state = {}
+    if init_func is not None:
+        init_func(state, **init_kwargs)
+
+    barrier.wait()
+
+    while True:
+        if pipe.poll(0.05):
+            obj = pipe.recv()
+            cmd = obj['cmd']
+            if cmd == 'job':
+                result = process_func(state, obj['data'])
+                pipe.send({'cmd':'result', 'data': result})
+            elif cmd == 'finalize':
+                break
+
+    if final_func is not None:
+        final_func(state)
+
+def run_sequence(data_list : List,
+                 process_func : Callable,
+                 init_func : Callable = None, init_kwargs : dict = None,
+                 final_func : Callable = None,
+                 mp_count : int = None, progress_bar_desc='Processing'):
+    """
+    Simple Job to process list of picklable data.
+
+     init_func(state:dict, **init_kwargs)
+
+     process_func(state:dict, data) -> object
+
+     mp_count(None)     number of subprocesses. Default - cores count.
+    """
+    if mp_count is None:
+        mp_count = multiprocessing.cpu_count()
+
+    barrier = multiprocessing.Barrier(mp_count)
+
+    n_data_sent = [0]*mp_count
+    conn_list = [None]*mp_count
+    p_list = [None]*mp_count
+
+    for i in range(mp_count):
+        s_pipe, c_pipe = conn_list[i] = multiprocessing.Pipe()
+        p = p_list[i] = multiprocessing.Process(target=_run_sequence, args=(barrier, init_func, init_kwargs, final_func, process_func, c_pipe), daemon=True )
+        p.start()
+
+    data_list_len = len(data_list)
+    n_data_done = 0
+    i_data = 0
+
+    lib_con.progress_bar_print(0, data_list_len, desc=progress_bar_desc)
+    result = []
+    while n_data_done != data_list_len:
+
+        for n_conn, (s_pipe, _) in enumerate(conn_list):
+
+            if i_data < data_list_len:
+                if n_data_sent[n_conn] < 2:
+                    n_data_sent[n_conn] += 1
+
+                    data = data_list[i_data]
+                    i_data += 1
+
+                    s_pipe.send( {'cmd':'job', 'data':data} )
+
+            if s_pipe.poll(0):
+                obj = s_pipe.recv()
+
+                cmd = obj['cmd']
+                if cmd == 'result':
+                    n_data_done += 1
+                    lib_con.progress_bar_print(n_data_done, data_list_len, desc=progress_bar_desc)
+
+                    n_data_sent[n_conn] -= 1
+
+                    data = obj['data']
+                    if data is not None:
+                        result.append(data)
+
+    for n_conn, (s_pipe, _) in enumerate(conn_list):
+        s_pipe.send( {'cmd':'finalize'} )
+
+    return result