diff --git a/omniadvisor/src/omniadvisor/interface/hijack_recommend.py b/omniadvisor/src/omniadvisor/interface/hijack_recommend.py
index cd0ab825628942c5be72cc35ee57656c3bcbb968..dd8ecba2d25aa2ac1d021bad6354d3f15d62cac0 100644
--- a/omniadvisor/src/omniadvisor/interface/hijack_recommend.py
+++ b/omniadvisor/src/omniadvisor/interface/hijack_recommend.py
@@ -260,6 +260,8 @@ def hijack_recommend(argv: list) -> None:
     :param argv: Spark任务的执行命令
     :return:
     """
+    # 对spark入参中的driver_java_options参数进行预处理
+    SparkCMDParser.pre_handler_driver_java_options(argv)
 
     # 非SUBMIT动作(指kill任务/查询状态/查询版本)的提交直接回退到原生spark-submit脚本执行 不被特性所劫持
     SparkCMDParser.validate_submit_arguments(argv)
diff --git a/omniadvisor/src/omniadvisor/service/spark_service/spark_cmd_parser.py b/omniadvisor/src/omniadvisor/service/spark_service/spark_cmd_parser.py
index 0143aa4ddb87ab8a0dfb9f41f0239e9315853839..47162de1042d261fcb646192b71d26010da9b4cf 100644
--- a/omniadvisor/src/omniadvisor/service/spark_service/spark_cmd_parser.py
+++ b/omniadvisor/src/omniadvisor/service/spark_service/spark_cmd_parser.py
@@ -14,6 +14,10 @@ _CMD_PREFIX_KEY = 'omniadvisor-cmd-prefix'
 _CMD_PRIMARY_RESOURCE = 'primary_resource'
 # 命令行中class字段
 _CMD_CLASS_KEY = 'class'
+# spark-sql命令在spark-submit中对应的的类名
+_SPARK_SQL_CLASS = 'org.apache.spark.sql.hive.thriftserver.SparkSQLCLIDriver'
+# spark参数driver-java-options
+_DRIVER_JAVA_OPTIONS = '--driver-java-options'
 # 单横杠参数的字段
 _SINGLE_HORIZONTAL_BAR_KEYS = ['e', 'f', 'i']
 # 布尔类型的字段
@@ -28,12 +32,18 @@ _SPARK_CONF_SUPPLEMENT_MAP = {
 }
 
 
+class SilentArgumentParser(argparse.ArgumentParser):
+    def error(self, message):
+        # 不打印错误 把异常往上层抛出
+        raise argparse.ArgumentError(None, message)
+
+
 class SparkCMDParser:
     """
     Spark 引擎的解析器
     """
     # 当用户提交-h和--help时需要退回原生执行 因此关闭argparse自带的--help/-h参数解析
-    _parser = argparse.ArgumentParser(add_help=False)
+    _parser = SilentArgumentParser(add_help=False)
 
     # 参数命中包含'-'的参数，设置dest属性，避免参数解析中'-'被转换为'_'
     # 解析Spark CMD开头，需脚本配合
@@ -88,8 +98,20 @@ class SparkCMDParser:
     _parser.add_argument('-f', type=str, help='File containing SQL script.')
     _parser.add_argument('-i', help='Initialization SQL file')
     _parser.add_argument('-d', '--define')
-    _parser.add_argument('--database')
+    _parser.add_argument('-database', '--database')
 
+    @classmethod
+    def pre_handler_driver_java_options(cls, argv):
+        """
+        :param argv: 对入参列表中的argv中driver-java-options的参数值进行处理 使用双引号包裹方便后续argparse模块识别
+        :return:
+        """
+        global_logger.debug(f"Before pre_handler_driver_java_options %s", argv)
+        driver_java_options_indices = [i for i, arg in enumerate(argv) if arg == _DRIVER_JAVA_OPTIONS]
+        for idx in driver_java_options_indices:
+            if idx + 1 < len(argv):
+                argv[idx + 1] = f'\"{argv[idx + 1]}\"'
+        global_logger.debug(f"After pre_handler_driver_java_options执行后 %s", argv)
 
     @classmethod
     def validate_submit_arguments(cls, args: List) -> None:
@@ -107,9 +129,12 @@ class SparkCMDParser:
         # 待检查参数列表
         params, unknown = cls._parser.parse_known_args(args=args)
         primary_resource = unknown
-
+        # 非spark-submit的提交类任务
         if not primary_resource and vars(params)['class'] is None:
-            raise TypeError("This is not a SUBMIT type spark task.")
+            raise TypeError("This is not a spark-submit SUBMIT type spark task.")
+        # 检查是否为spark-sql提交的执行类任务
+        if vars(params)['class'] == _SPARK_SQL_CLASS and vars(params)['e'] is None and vars(params)['f'] is None:
+            raise TypeError("This is not a spark-sql SUBMIT type spark task.")
 
     @staticmethod
     def _normalize_value(value: Any) -> Any: