diff --git a/conf-template/client_import/csv_datasource.conf b/conf-template/client_import/csv_datasource.conf
new file mode 100644
index 00000000..9936bdda
--- /dev/null
+++ b/conf-template/client_import/csv_datasource.conf
@@ -0,0 +1,94 @@
+# Use the command to submit the exchange job:
+
+# spark-submit \
+# --master "spark://master_ip:7077" \
+# --driver-memory=2G --executor-memory=30G  \
+# --num-executors=3 --executor-cores=20 \
+# --class com.vesoft.nebula.exchange.Exchange \
+# nebula-exchange-3.0-SNAPSHOT.jar -c csv_datasource.conf
+
+{
+  # Spark config
+  spark: {
+    app: {
+      name: NebulaGraph Exchange
+    }
+  }
+
+  # Nebula Graph config
+  nebula: {
+    address:{
+      graph:["127.0.0.1:9669"]
+      # if your NebulaGraph server is in virtual network like k8s, please config the leader address of meta.
+      # use `SHOW meta leader` to see your meta leader's address
+      meta:["127.0.0.1:9559"]
+    }
+    user: root
+    pswd: nebula
+    space: test
+
+    # nebula client connection parameters
+    connection {
+      # socket connect & execute timeout, unit: millisecond
+      timeout: 30000
+    }
+
+    error: {
+      # max number of failures, if the number of failures is bigger than max, then exit the application.
+      max: 32
+      # failed data will be recorded in output path, format with ngql
+      output: /tmp/errors
+    }
+
+    # use google's RateLimiter to limit the requests send to NebulaGraph
+    rate: {
+      # the stable throughput of RateLimiter
+      limit: 1024
+      # Acquires a permit from RateLimiter, unit: MILLISECONDS
+      # if it can't be obtained within the specified timeout, then give up the request.
+      timeout: 1000
+    }
+  }
+
+  # Processing tags
+  tags: [
+    {
+      name: tag-name-1
+      type: {
+        source: csv
+        sink: client
+      }
+      # if your file in not in hdfs, config "file:///path/test.csv"
+      path: "hdfs://ip:port/path/test.csv"
+      # if your csv file has no header, then use _c0,_c1,_c2,.. to indicate fields
+      fields: [csv-field-0, csv-field-1, csv-field-2]
+      nebula.fields: [nebula-field-0, nebula-field-1, nebula-field-2]
+      vertex: csv-field-0
+      separator: ","
+      header: true
+      batch: 2000
+      partition: 60
+    }
+  ]
+
+  # process edges
+  edges: [
+    {
+      name: edge-name-1
+      type: {
+        source: csv
+        sink: client
+      }
+      path: "hdfs://ip:port/path/test.csv"
+      fields: [csv-field-0, csv-field-1, csv-field-2]
+      nebula.fields: [nebula-field-0, nebula-field-1, nebula-field-2]
+      source: csv-field-0
+      target: csv-field-1
+      ranking: csv-field-2
+      separator: ","
+      header: true
+      batch: 2000
+      partition: 60
+    }
+  ]
+}
diff --git a/conf-template/client_import/hive_datasource.conf b/conf-template/client_import/hive_datasource.conf
new file mode 100644
index 00000000..2b2c6e26
--- /dev/null
+++ b/conf-template/client_import/hive_datasource.conf
@@ -0,0 +1,88 @@
+# Use the command to submit the exchange job:
+
+# spark-submit \
+# --master "spark://master_ip:7077" \
+# --driver-memory=2G --executor-memory=30G  \
+# --num-executors=3 --executor-cores=20 \
+# --class com.vesoft.nebula.exchange.Exchange \
+# nebula-exchange-3.0-SNAPSHOT.jar -c csv_datasource.conf
+
+{
+  # Spark config
+  spark: {
+    app: {
+      name: NebulaGraph Exchange
+    }
+  }
+
+  # Nebula Graph config
+  nebula: {
+    address:{
+      graph:["127.0.0.1:9669"]
+      # if your NebulaGraph server is in virtual network like k8s, please config the leader address of meta.
+      # use `SHOW meta leader` to see your meta leader's address
+      meta:["127.0.0.1:9559"]
+    }
+    user: root
+    pswd: nebula
+    space: test
+
+    # nebula client connection parameters
+    connection {
+      # socket connect & execute timeout, unit: millisecond
+      timeout: 30000
+    }
+
+    error: {
+      # max number of failures, if the number of failures is bigger than max, then exit the application.
+      max: 32
+      # failed data will be recorded in output path, format with ngql
+      output: /tmp/errors
+    }
+
+    # use google's RateLimiter to limit the requests send to NebulaGraph
+    rate: {
+      # the stable throughput of RateLimiter
+      limit: 1024
+      # Acquires a permit from RateLimiter, unit: MILLISECONDS
+      # if it can't be obtained within the specified timeout, then give up the request.
+      timeout: 1000
+    }
+  }
+
+  # Processing tags
+  tags: [
+    {
+      name: tag-name-1
+      type: {
+        source: hive
+        sink: client
+      }
+      exec: "select hive-field0, hive-field1, hive-field2 from database.table"
+      fields: [hive-field-0, hive-field-1, hive-field-2]
+      nebula.fields: [nebula-field-0, nebula-field-1, nebula-field-2]
+      vertex: hive-field-0
+      batch: 2000
+      partition: 60
+    }
+  ]
+
+  # process edges
+  edges: [
+    {
+      name: edge-name-1
+      type: {
+        source: hive
+        sink: client
+      }
+      exec: "select hive-field0, hive-field1, hive-field2 from database.table"
+      fields: [ hive-field-0, hive-field-1, hive-field-2]
+      nebula.fields: [nebula-field-0, nebula-field-1, nebula-field-2]
+      source: hive-field-0
+      target: hive-field-1
+      ranking: hive-filed-2
+      batch: 2000
+      partition: 60
+    }
+  ]
+}
diff --git a/conf-template/sst_import/csv_datasource.conf b/conf-template/sst_import/csv_datasource.conf
new file mode 100644
index 00000000..dd20466d
--- /dev/null
+++ b/conf-template/sst_import/csv_datasource.conf
@@ -0,0 +1,101 @@
+# Use the command to submit the exchange job:
+
+# spark-submit \
+# --master "spark://master_ip:7077" \
+# --driver-memory=2G --executor-memory=30G  \
+# --num-executors=3 --executor-cores=20 \
+# --class com.vesoft.nebula.exchange.Exchange \
+# nebula-exchange-3.0-SNAPSHOT.jar -c csv_datasource.conf
+
+{
+  # Spark config
+  spark: {
+    app: {
+      name: NebulaGraph Exchange
+    }
+  }
+
+  # Nebula Graph config
+  nebula: {
+    address:{
+      graph:["127.0.0.1:9669"]
+      # if your NebulaGraph server is in virtual network like k8s, please config the leader address of meta.
+      # use `SHOW meta leader` to see your meta leader's address
+      meta:["127.0.0.1:9559"]
+    }
+    user: root
+    pswd: nebula
+    space: test
+
+    path:{
+        # any path that owns read and write access is ok
+        local:"/tmp"
+        remote:"/sst"
+        hdfs.namenode: "hdfs://name_node:9000"
+    }
+
+    # nebula client connection parameters
+    connection {
+      # socket connect & execute timeout, unit: millisecond
+      timeout: 30000
+    }
+
+    error: {
+      # max number of failures, if the number of failures is bigger than max, then exit the application.
+      max: 32
+      # failed data will be recorded in output path, format with ngql
+      output: /tmp/errors
+    }
+
+    # use google's RateLimiter to limit the requests send to NebulaGraph
+    rate: {
+      # the stable throughput of RateLimiter
+      limit: 1024
+      # Acquires a permit from RateLimiter, unit: MILLISECONDS
+      # if it can't be obtained within the specified timeout, then give up the request.
+      timeout: 1000
+    }
+  }
+
+  # Processing tags
+  tags: [
+    {
+      name: tag-name-1
+      type: {
+        source: csv
+        sink: sst
+      }
+      # if your file in not in hdfs, config "file:///path/test.csv"
+      path: "hdfs://ip:port/path/test.csv"
+      # if your csv file has no header, then use _c0,_c1,_c2,.. to indicate fields
+      fields: [csv-field-0, csv-field-1, csv-field-2]
+      nebula.fields: [nebula-field-0, nebula-field-1, nebula-field-2]
+      vertex: csv-field-0
+      separator: ","
+      header: true
+      batch: 2000
+      partition: 60
+    }
+  ]
+
+  # process edges
+  edges: [
+    {
+      name: edge-name-1
+      type: {
+        source: csv
+        sink: sst
+      }
+      path: "hdfs://ip:port/path/test.csv"
+      fields: [csv-field-0, csv-field-1, csv-field-2]
+      nebula.fields: [nebula-field-0, nebula-field-1, nebula-field-2]
+      source: csv-field-0
+      target: csv-field-1
+      ranking: csv-field-2
+      separator: ","
+      header: true
+      batch: 2000
+      partition: 60
+    }
+  ]
+}
diff --git a/conf-template/sst_import/hive_datasource.conf b/conf-template/sst_import/hive_datasource.conf
new file mode 100644
index 00000000..fdbbc0e0
--- /dev/null
+++ b/conf-template/sst_import/hive_datasource.conf
@@ -0,0 +1,95 @@
+# Use the command to submit the exchange job:
+
+# spark-submit \
+# --master "spark://master_ip:7077" \
+# --driver-memory=2G --executor-memory=30G  \
+# --num-executors=3 --executor-cores=20 \
+# --class com.vesoft.nebula.exchange.Exchange \
+# nebula-exchange-3.0-SNAPSHOT.jar -c csv_datasource.conf
+
+{
+  # Spark config
+  spark: {
+    app: {
+      name: NebulaGraph Exchange
+    }
+  }
+
+  # Nebula Graph config
+  nebula: {
+    address:{
+      graph:["127.0.0.1:9669"]
+      # if your NebulaGraph server is in virtual network like k8s, please config the leader address of meta.
+      # use `SHOW meta leader` to see your meta leader's address
+      meta:["127.0.0.1:9559"]
+    }
+    user: root
+    pswd: nebula
+    space: test
+
+    path:{
+        # any path that owns read and write access is ok
+        local:"/tmp"
+        remote:"/sst"
+        hdfs.namenode: "hdfs://name_node:9000"
+    }
+
+    # nebula client connection parameters
+    connection {
+      # socket connect & execute timeout, unit: millisecond
+      timeout: 30000
+    }
+
+    error: {
+      # max number of failures, if the number of failures is bigger than max, then exit the application.
+      max: 32
+      # failed data will be recorded in output path, format with ngql
+      output: /tmp/errors
+    }
+
+    # use google's RateLimiter to limit the requests send to NebulaGraph
+    rate: {
+      # the stable throughput of RateLimiter
+      limit: 1024
+      # Acquires a permit from RateLimiter, unit: MILLISECONDS
+      # if it can't be obtained within the specified timeout, then give up the request.
+      timeout: 1000
+    }
+  }
+
+  # Processing tags
+  tags: [
+    {
+      name: tag-name-1
+      type: {
+        source: hive
+        sink: sst
+      }
+      exec: "select hive-field0, hive-field1, hive-field2 from database.table"
+      fields: [hive-field-0, hive-field-1, hive-field-2]
+      nebula.fields: [nebula-field-0, nebula-field-1, nebula-field-2]
+      vertex: hive-field-0
+      batch: 2000
+      partition: 60
+    }
+  ]
+
+  # process edges
+  edges: [
+    {
+      name: edge-name-1
+      type: {
+        source: hive
+        sink: sst
+      }
+      exec: "select hive-field0, hive-field1, hive-field2 from database.table"
+      fields: [ hive-field-0, hive-field-1, hive-field-2]
+      nebula.fields: [nebula-field-0, nebula-field-1, nebula-field-2]
+      source: hive-field-0
+      target: hive-field-1
+      ranking: hive-filed-2
+      batch: 2000
+      partition: 60
+    }
+  ]
+}
diff --git a/nebula-exchange_spark_2.4/src/main/resources/application.conf b/nebula-exchange_spark_2.4/src/main/resources/application.conf
index c3028227..131f0cf7 100644
--- a/nebula-exchange_spark_2.4/src/main/resources/application.conf
+++ b/nebula-exchange_spark_2.4/src/main/resources/application.conf
@@ -1,23 +1,8 @@
 {
-  # Spark relation com.vesoft.exchange.common.config
+  # Spark config
   spark: {
     app: {
-      name: Nebula Exchange 2.0
-    }
-
-    master:local
-
-    driver: {
-      cores: 1
-      maxResultSize: 1G
-    }
-
-    executor: {
-        memory:1G
-    }
-
-    cores:{
-      max: 16
+      name: Nebula Exchange
     }
   }
 
@@ -33,7 +18,7 @@
   #  }
 
 
-  # Nebula Graph relation com.vesoft.exchange.common.config
+  # Nebula Graph config
   nebula: {
     address:{
       graph:["127.0.0.1:9669"]
@@ -118,8 +103,8 @@
         field:parquet-field-0
         #policy:hash
       }
-      batch: 256
-      partition: 32
+      batch: 2000
+      partition: 60
     }
 
     # HDFS csv
@@ -139,8 +124,8 @@
       }
       separator: ","
       header: true
-      batch: 256
-      partition: 32
+      batch: 2000
+      partition: 60
       # optional config, default is false
       # config repartitionWithNebula as true means: repartition spark dataframe with nebula partition number to write sst files.
       repartitionWithNebula: false
@@ -160,8 +145,8 @@
         field: json-field-0
         #policy: hash
       }
-      batch: 256
-      partition: 32
+      batch: 2000
+      partition: 60
     }
 
     # Hive
@@ -178,8 +163,8 @@
         field: hive-field-0
         # policy: "hash"
       }
-      batch: 256
-      partition: 32
+      batch: 2000
+      partition: 60
     }
 
     # neo4j
@@ -199,8 +184,8 @@
         field:neo4j-field-0
         # policy:hash
       }
-      partition: 10
-      batch: 1000
+      partition: 60
+      batch: 2000
       check_point_path: /tmp/test
    }
 
@@ -221,8 +206,8 @@
       vertex: {
         field:rowkey
       }
-      partition: 10
-      batch: 1000
+      partition: 60
+      batch: 2000
     }
 
     # Pulsar
@@ -243,8 +228,8 @@
       vertex: {
         field:pulsar-field-0
       }
-      partition: 10
-      batch: 1000
+      partition: 60
+      batch: 2000
       interval.seconds: 10
     }
 
@@ -262,8 +247,8 @@
       vertex: {
         field: kafka-field-0
       }
-      partition: 10
-      batch: 10
+      partition: 60
+      batch: 2000
       interval.seconds: 10
     }
 
@@ -290,8 +275,8 @@
       vertex:{
         field: maxcompute-field-2
       }
-      partition:10
-      batch:10
+      partition:60
+      batch:2000
     }
 
     # ClickHouse
@@ -313,8 +298,8 @@
         field:clickhouse-field-0
         #policy:hash
       }
-      batch: 256
-      partition: 32
+      batch: 2000
+      partition: 60
     }
 
     # PostgreSQL
@@ -338,8 +323,8 @@
         field: postgre-field-0
         # policy: "hash"
       }
-      batch: 256
-      partition: 32
+      batch: 2000
+      partition: 60
     }
 
     # Oracle
@@ -361,8 +346,8 @@
         field: oracle-field-0
         # policy: "hash"
       }
-      batch: 256
-      partition: 32
+      batch: 2000
+      partition: 60
     }
   ]
 
@@ -389,8 +374,8 @@
         #policy:hash
       }
       ranking: parquet-field-2
-      batch: 256
-      partition: 32
+      batch: 2000
+      partition: 60
     }
 
     # HDFS csv
@@ -413,8 +398,8 @@
       ranking: csv-field-2
       separator: ","
       header: true
-      batch: 256
-      partition: 32
+      batch: 2000
+      partition: 60
       # optional config, default is false
       # config repartitionWithNebula as true means: repartition spark dataframe with nebula partition number to write sst files.
       repartitionWithNebula: false
@@ -438,8 +423,8 @@
         field: json-field-1
       }
       ranking: json-field-2
-      batch: 256
-      partition: 32
+      batch: 2000
+      partition: 60
     }
 
     # Hive
@@ -455,8 +440,8 @@
       source: hive-field-0
       target: hive-field-1
       ranking: hive-filed-2
-      batch: 256
-      partition: 32
+      batch: 2000
+      partition: 60
     }
 
     # Neo4j
@@ -479,8 +464,8 @@
         field: b.neo4j-target-field
       }
       ranking: neo4j-field-2
-      partition: 10
-      batch: 1000
+      partition: 60
+      batch: 2000
       check_point_path: /tmp/test
     }
 
@@ -504,8 +489,8 @@
         field: hbase-column-h
       }
       ranking: hbase-column-t
-      partition: 10
-      batch: 1000
+      partition: 60
+      batch: 2000
     }
 
 
@@ -532,8 +517,8 @@
         field: pulsar-field-1
       }
       ranking: pulsar-field-2
-      partition: 10
-      batch: 10
+      partition: 60
+      batch: 2000
       interval.seconds: 10
     }
 
@@ -551,8 +536,8 @@
       source: kafka-field-0
       target: kafka-field-1
       ranking: kafka-field-2
-      partition: 10
-      batch: 1000
+      partition: 60
+      batch: 2000
       interval.seconds: 10
     }
 
@@ -581,8 +566,8 @@
         field: maxcompute-field-3
       }
       ranking: maxcompute-field-4
-      partition:10
-      batch:10
+      partition:60
+      batch:2000
     }
 
     # ClickHouse
@@ -609,8 +594,8 @@
         #policy:hash
       }
       ranking:clickhouse-field-3
-      batch: 256
-      partition: 32
+      batch: 2000
+      partition: 60
     }
 
     # PostgreSQL
@@ -639,8 +624,8 @@
         # policy: "hash"
       }
       ranking: postgre-field-1
-      batch: 256
-      partition: 32
+      batch: 2000
+      partition: 60
     }
 
     # Oracle
@@ -666,8 +651,8 @@
         field: oracle-field-1
       }
       ranking: oracle-field-2
-      batch: 256
-      partition: 32
+      batch: 2000
+      partition: 60
     }
   ]
 }