diff --git a/src/observer/sql/parser/lex_sql.l b/src/observer/sql/parser/lex_sql.l
index ed028d048..26727a30b 100644
--- a/src/observer/sql/parser/lex_sql.l
+++ b/src/observer/sql/parser/lex_sql.l
@@ -27,6 +27,7 @@ while (0);
 %{
 #include<string.h>
 #include<stdio.h>
+#include <vector>
 
 /**
  * flex 代码包含三个部分，使用 %% 分隔
@@ -119,7 +120,7 @@ GROUP                                   RETURN_TOKEN(GROUP);
 BY                                      RETURN_TOKEN(BY);
 STORAGE                                 RETURN_TOKEN(STORAGE);
 FORMAT                                  RETURN_TOKEN(FORMAT);
-{ID}                                    yylval->cstring=strdup(yytext); RETURN_TOKEN(ID);
+{ID}                                    yylval->cstring=strdup(yytext); static_cast<std::vector<char*>*>(yyextra)->push_back(yylval->cstring); RETURN_TOKEN(ID);
 "("                                     RETURN_TOKEN(LBRACE);
 ")"                                     RETURN_TOKEN(RBRACE);
 
@@ -136,8 +137,8 @@ FORMAT                                  RETURN_TOKEN(FORMAT);
 "-" |
 "*" |
 "/"                                     { return yytext[0]; }
-\"[^"]*\"                               yylval->cstring = strdup(yytext); RETURN_TOKEN(SSS);
-'[^']*\'                                yylval->cstring = strdup(yytext); RETURN_TOKEN(SSS);
+\"[^"]*\"                               yylval->cstring = strdup(yytext); static_cast<std::vector<char*>*>(yyextra)->push_back(yylval->cstring); RETURN_TOKEN(SSS);
+'[^']*\'                                yylval->cstring = strdup(yytext); static_cast<std::vector<char*>*>(yyextra)->push_back(yylval->cstring); RETURN_TOKEN(SSS);
 
 .                                       LOG_DEBUG("Unknown character [%c]",yytext[0]); return yytext[0];
 %%
diff --git a/src/observer/sql/parser/yacc_sql.y b/src/observer/sql/parser/yacc_sql.y
index 786879000..db210e970 100644
--- a/src/observer/sql/parser/yacc_sql.y
+++ b/src/observer/sql/parser/yacc_sql.y
@@ -254,7 +254,6 @@ drop_table_stmt:    /*drop table 语句的语法解析树*/
     DROP TABLE ID {
       $$ = new ParsedSqlNode(SCF_DROP_TABLE);
       $$->drop_table.relation_name = $3;
-      free($3);
     };
 
 show_tables_stmt:
@@ -267,7 +266,6 @@ desc_table_stmt:
     DESC ID  {
       $$ = new ParsedSqlNode(SCF_DESC_TABLE);
       $$->desc_table.relation_name = $2;
-      free($2);
     }
     ;
 
@@ -279,9 +277,6 @@ create_index_stmt:    /*create index 语句的语法解析树*/
       create_index.index_name = $3;
       create_index.relation_name = $5;
       create_index.attribute_name = $7;
-      free($3);
-      free($5);
-      free($7);
     }
     ;
 
@@ -291,8 +286,6 @@ drop_index_stmt:      /*drop index 语句的语法解析树*/
       $$ = new ParsedSqlNode(SCF_DROP_INDEX);
       $$->drop_index.index_name = $3;
       $$->drop_index.relation_name = $5;
-      free($3);
-      free($5);
     }
     ;
 create_table_stmt:    /*create table 语句的语法解析树*/
@@ -301,7 +294,7 @@ create_table_stmt:    /*create table 语句的语法解析树*/
       $$ = new ParsedSqlNode(SCF_CREATE_TABLE);
       CreateTableSqlNode &create_table = $$->create_table;
       create_table.relation_name = $3;
-      free($3);
+      //free($3);
 
       vector<AttrInfoSqlNode> *src_attrs = $6;
 
@@ -314,7 +307,6 @@ create_table_stmt:    /*create table 语句的语法解析树*/
       delete $5;
       if ($8 != nullptr) {
         create_table.storage_format = $8;
-        free($8);
       }
     }
     ;
@@ -342,7 +334,6 @@ attr_def:
       $$->type = (AttrType)$2;
       $$->name = $1;
       $$->length = $4;
-      free($1);
     }
     | ID type
     {
@@ -350,7 +341,6 @@ attr_def:
       $$->type = (AttrType)$2;
       $$->name = $1;
       $$->length = 4;
-      free($1);
     }
     ;
 number:
@@ -374,7 +364,6 @@ insert_stmt:        /*insert   语句的语法解析树*/
       $$->insertion.values.emplace_back(*$6);
       reverse($$->insertion.values.begin(), $$->insertion.values.end());
       delete $6;
-      free($3);
     }
     ;
 
@@ -406,7 +395,6 @@ value:
       char *tmp = common::substr($1,1,strlen($1)-2);
       $$ = new Value(tmp);
       free(tmp);
-      free($1);
     }
     ;
 storage_format:
@@ -429,7 +417,6 @@ delete_stmt:    /*  delete 语句的语法解析树*/
         $$->deletion.conditions.swap(*$4);
         delete $4;
       }
-      free($3);
     }
     ;
 update_stmt:      /*  update 语句的语法解析树*/
@@ -443,8 +430,6 @@ update_stmt:      /*  update 语句的语法解析树*/
         $$->update.conditions.swap(*$7);
         delete $7;
       }
-      free($2);
-      free($4);
     }
     ;
 select_stmt:        /*  select 语句的语法解析树*/
@@ -538,14 +523,11 @@ rel_attr:
     ID {
       $$ = new RelAttrSqlNode;
       $$->attribute_name = $1;
-      free($1);
     }
     | ID DOT ID {
       $$ = new RelAttrSqlNode;
       $$->relation_name  = $1;
       $$->attribute_name = $3;
-      free($1);
-      free($3);
     }
     ;
 
@@ -558,7 +540,6 @@ rel_list:
     relation {
       $$ = new vector<string>();
       $$->push_back($1);
-      free($1);
     }
     | relation COMMA rel_list {
       if ($3 != nullptr) {
@@ -568,7 +549,6 @@ rel_list:
       }
 
       $$->insert($$->begin(), $1);
-      free($1);
     }
     ;
 
@@ -672,7 +652,6 @@ load_data_stmt:
       $$ = new ParsedSqlNode(SCF_LOAD_DATA);
       $$->load_data.relation_name = $7;
       $$->load_data.file_name = tmp_file_name;
-      free($7);
       free(tmp_file_name);
     }
     ;
@@ -691,7 +670,6 @@ set_variable_stmt:
       $$ = new ParsedSqlNode(SCF_SET_VARIABLE);
       $$->set_variable.name  = $2;
       $$->set_variable.value = *$4;
-      free($2);
       delete $4;
     }
     ;
@@ -705,9 +683,16 @@ extern void scan_string(const char *str, yyscan_t scanner);
 
 int sql_parse(const char *s, ParsedSqlResult *sql_result) {
   yyscan_t scanner;
-  yylex_init(&scanner);
+  std::vector<char *> allocated_strings;
+  yylex_init_extra(static_cast<void*>(&allocated_strings),&scanner);
   scan_string(s, scanner);
   int result = yyparse(s, sql_result, scanner);
+
+  for (char *ptr : allocated_strings) {
+    free(ptr);
+  }
+  allocated_strings.clear();
+
   yylex_destroy(scanner);
   return result;
-}
+}
\ No newline at end of file