diff --git a/doc/source/ch08_csst_common.rst b/doc/source/ch08_csst_common.rst
index 1ed79cde1b2df1cc79933fbd011c8b1493741173..c14cd0066074c931eb23526034d33b1f47079470 100644
--- a/doc/source/ch08_csst_common.rst
+++ b/doc/source/ch08_csst_common.rst
@@ -17,32 +17,51 @@ To initilize from local directory:
 
     from csst_common.data_manager import CsstMsDataManager
     # for full basic initialization
-    dm_mbi = CsstMsDataManager.from_dir(
+    dm = CsstMsDataManager.from_dir(
         ver_sim="C5.2",                  # version of simulation, "C5.2" is the latest
-        dir_l0="",                       # the L0/input directory
-        dir_l1="",                       # the L1/output directory
-        dir_pcref="",                    # position calibration reference (will be deprecated)
+        datatype="mbi",                  # data type, "mbi", "sls" or "all"
+        dir_l0=".",                      # the L0/input directory
+        dir_l1=".",                      # the L1/output directory
         path_aux="",                     # aux file paths (master bias, dark, flat)
-        datatype="mbi",                  # "mbi", "sls" or "all"
+        dfs_mode=False,                  # if True, use DFS in pipeline modules
+        dfs_node='kmust',                # DFS server, chosen from {"kmust", "pml"}
     )
-    dm_mbi.set_detectors(detectors=None) # use all available detectors
+    dm.target_detectors = None           # use all available detectors
 
-In particular, if you are on ``dandelion`` or ``PM node``, you can use a ``quickstart`` method.
+
+In particular, if you are on ``dandelion`` or ``PML node``, you can use a ``quickstart`` method.
 
 .. code-block:: python
 
     # for a quick start ()
-    dm_mbi = CsstMsDataManager.quickstart(
-        ver_sim="C5.2",
-        datatype="mbi",
-        dir_l1=".",
-        exposure_id=100                  # the 100th exposure
+    dm = CsstMsDataManager.quickstart(
+        ver_sim='C5.2',                  # version of simulation, only "C5.2" is supported
+        datatype='mbi',                  # data type, "mbi", "sls" or "all"
+        dir_l1='.',                      # the L1/output directory
+        exposure_id=100,                 # the exposure ID
+        dfs_mode=False,                  # if True, use DFS in pipeline modules
+        dfs_node='kmust',                # DFS server, chosen from {"kmust", "pml"}
     )
-    dm_mbi.set_detectors(detectors=None) # use all available detectors
+    dm
+
+You will get the outputs:
+
+.. code-block::
+
+    <CsstMsDataManager>
+    Data type = mbi
+    Valid detectors = [6, 7, 8, 9, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 22, 23, 24, 25]
+    Available detectors = [6, 7, 8, 9, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 22, 23, 24, 25]
+    Target detectors = [6, 7, 8, 9, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 22, 23, 24, 25]
+    dir_l0 = /nfsdata/share/csst_simulation_data/Cycle-5-SimuData/multipleBandsImaging/NGP_AstrometryON_shearOFF/MSC_0000100/
+    dir_l1 = .
+    dfs_mode = False
+    dfs_node = kmust
+
 
 .. note::
     This ``quickstart`` method is ONLY available on
-    ``dandelion`` or the ``Purple Mountain Lab node``.
+    ``dandelion`` or the ``PML node``.
     In your local development environment, you have to use
     ``CsstMsDataManager.from_dir(...)`` to initialize it.