达梦参数文件dm.ini

达梦部署 专栏收录该内容
7 篇文章 1 订阅

查看参数可以使用v d m i n i 或 v dm_ini或v dminivparameter视图
SF_GET_PARA_VALUE、SF_GET_PARA_DOUBLE_VALUE、SF_GET_PARA_STRING_VALUE。

select distinct para_type from v$dm_ini;
1、READ ONLY:手动参数,表示服务器运行过程中不可修改;
2、IN FILE:静态参数,只可修改ini文件;
3、SYS和SESSION:动态参数,ini文件和内存同时可修改,其中,SYS系统级参数;

静态:可以被动态修改,修改后重启服务器才能生效。
动态:可以被动态修改,修改后即时生效。动态参数又分为会话级和系统级两种。会话级参数被修改后,新参数值只会影响新创建的会话,之前创建的会话不受影响;系统级参数的修改则会影响所有的会话。
手动:不能被动态修改,必须手动修改dm.ini参数文件,然后重启才能生效。

一、修改参数的方法:
1、SP_SET_PARA_VALUE (scope int, paraname varchar(256), value int64)
该过程用于修改整型静态配置参数和动态配置参数。SCOPE参数为1表示在内存和INI文件中都修改参数值,此时只能修改动态的配置参数。参数为2表示只在INI文件中修改配置参数,此时可用来修改静态配置参数和动态配置参数
2、SP_SET_PARA_DOUBLE_VALUE (scope int, paraname varchar(8187), value double)
该过程用于修改浮点型静态配置参数和动态配置参数。SCOPE参数为1表示在内存和INI文件中都修改参数值,此时只能修改动态的配置参数。参数为2表示只在INI文件中修改配置参数,此时可用来修改静态配置参数和动态配置参数
3、SP_SET_PARA_STRING_VALUE()
该过程用于修改字符串型静态配置参数和动态配置参数。SCOPE参数为1表示在内存和INI文件中都修改参数值,此时只能修改动态的配置参数。参数为2表示只在INI文件中修改配置参数,此时可用来修改静态配置参数和动态配置参数
4、SF_SET_SYSTEM_PARA_VALUE (paraname varchar(256), value int64\double\varchar(256), deferred int, scope int64)
该过程用于修改系统整型、double、varchar的静态配置参数或动态配置参数。DEFERRED参数,为0表示当前session修改的参数立即生效,为1表示当前session不生效,后续再生效,默认为0。SCOPE参数为1表示在内存和INI文件中都修改参数值,此时只能修改动态的配置参数。参数为2表示只在INI文件中修改配置参数,此时可用来修改静态配置参数和动态配置参数。
5、SF_SET_SESSION_PARA_VALUE (paraname varchar(8187), value bigint)
设置某个会话级INI参数的值,设置的参数值只对本会话有效。
6、ALTER SYSTEM SET ‘参数名称’ =‘参数值’ [DEFERRED] [MEMORY|BOTH|SPFILE];
参数名称:可以是静态参数、系统参数sys、会话参数session;DEFERRED只能用于动态参数。指定DEFERRED后,参数值延迟生效,对当前session不生效,只对新创建的会话生效。默认为立即生效,对当前会话和新创建的会话都生效。[MEMORY|BOTH|SPFILE] 指定生效的位置。其中,MEMORY只对内存中的值做修改;SPFILE则只对file_value中的值做修改;BOTH则都做修改。默认情况下,为MEMORY。对于静态参数,只能指定SPFILE。
7、ALTER SESSION SET ‘参数名称’ =‘参数值’ [PURGE];
参数名称指动态会话级参数名字.[PURGE] 指是否清理执行计划。设置后的值只对当前会话有效。当包含 PURGE 选项时会清除服务器保存的所有执行计划。

1、READ ONLY:
手动编辑dm.ini

2、INFILE :
修改后重启生效,使用SP_SET_PARA_VALUE(修改整型,scope选择2)、SP_SET_PARA_STRING_VALUE(修改字符串型,scope选择2)
也可以使用ALTER SYSTEM SET ‘参数名称’ =‘参数值’ [DEFERRED] [MEMORY|BOTH|SPFILE];
SP_SET_PARA_VALUE (2, ‘LOG_POOL_SIZE’,512);
alter system set ‘LOG_POOL_SIZE’=256 SPFILE;

3、session :SF_SET_SESSION_PARA_VALUE(修改当前session,其他session无效),SP_SET_PARA_VALUE、SP_SET_PARA_DOUBLE_VALUE、SP_SET_PARA_STRING_VALUE(),修改session级别的参数不能对以前连接的session生效,可以选择以后的连接以及重启数据库后是否生效。
或ALTER SESSION SET ‘参数名称’ =‘参数值’ [PURGE];

SF_SET_SESSION_PARA_VALUE(‘SORT_BUF_SIZE’,100);
SP_SET_PARA_VALUE(1,‘SORT_BUF_SIZE’,400);
alter session set ‘SORT_BUF_SIZE’=50;

4、SYS: 动态参数,可以使用SF_SET_SYSTEM_PARA_VALUE函数修改,这样不用考虑整型、浮点、字符串的问题(当然也可以使用其他函数包)
或使用ALTER SYSTEM SET ‘参数名称’ =‘参数值’ [DEFERRED] [MEMORY|BOTH|SPFILE];

SF_SET_SYSTEM_PARA_VALUE (‘TIMER_TRIG_CHECK_INTERVAL’,30,0,1);
ALTER SYSTEM SET ‘TIMER_TRIG_CHECK_INTERVAL’ =60 DEFERRED BOTH;

READ ONLY和IN FILE的参数完全可以直接编辑dm.ini文件进行修改,然后重启数据库生效。这样就不用考虑应该使用哪个函数包了。
对于sys的参数,可以使用SF_SET_SESSION_PARA_VALUE函数包,避免考虑整型、浮点、字符型的问题。
对于session的参数一般使用SF_SET_SESSION_PARA_VALUE函数来修改当前session。
大多数情况下,使用SP_SET_PARA_VALUE就可以应对日常的运维工作了。
当然习惯兼容ORACLE运维的同学来说,使用alter system/session 也是个不错的选择。

控制文件相关参数

CTL_PATH 控制文件路径
CTL_BAK_PATH 控制文件备份路径
CTL_BAK_NUM 控制文件备份个数限制
SYSTEM_PATH 系统路径
CONFIG_PATH 指定DMSERVER所读取的配置文件(DMMAL.INI, DMARCH.INI, DMTIMER.INI等)的路径
TEMP_PATH 临时库文件路径
BAK_PATH 备份路径
DFS_PATH DFS_PATH

实例名

INSTANCE_NAME 实例名(一般情况下,长度不超过128个字符;但是在守护进程中,长度不超过16个字符)

内存相关参数

MAX_OS_MEMORY
100
DM服务器能使用的最大内存占操作系统的百分…当取值100时,服务器不进行内存的检查。

MEMORY_POOL
2000
系统内存池大小,以兆为单位。系统内存池是由DM管理的内存,系统内存的申请都从内存池中申请。有效值范围(5~10000)

MEMORY_TARGET
4000
共享内存池能扩充到的最大大小,以M为单位,有效值范围:32位平台为(02000),64位平台为(067108864),0表示不限制

MEMORY_EXTENT_SIZE
1
共享内存池每次扩充的大小,以M为单位,有效值范围(1~10240)

MEMORY_LEAK_CHECK
0
是否开启对所有内存池的泄漏检查,0:不开启;1:开启

MEMORY_MAGIC_CHECK
2
是否开启对所有内存池的校验,0:不开启;1:开启

MEMORY_BAK_POOL
4
系统备份内存池大小,以兆为单位。系统备份内存DM管理的内存。有效值范围(2~10000)

HUGE_MEMO…RCENTAGE
50
指示HUGE_BUFFER中可以借用作常规内存分配的空间百分比,有效值为0~100

HUGE_BUFFER
80
HFS表使用的缓冲区大小,以兆为单位。有效值范围(8~1048576)

HUGE_BUFFER_POOLS
4
number of Huge buffer pools

BUFFER
1000
系统缓冲区大小,以兆为单位。推荐值:系统缓冲区…%80%。有效值范围(81048576)

BUFFER_POOLS
43
BUFFER系统分区数(1~512)

FAST_POOL_PAGES
3000
快速缓冲区页面数(0~99999)

FAST_ROLL_PAGES
1000
快速缓冲区回滚页面数(0~99999)

KEEP
64
KEEP缓冲区大小,以兆为单位。有效值范围(8~1048576)

RECYCLE
1000
RECYCLE爱冲区大小,以兆为单位。有效值范围(8~1048576)

RECYCLE_POOLS
17
RECYCLE缓冲区分区数,每个RECYCLE分区的大小…/RECYCLE_POOLS。有效值范围(1~512)

ROLLSEG
1
system ROLLSEG buffer size in Megabytes

ROLLSEG_POOLS
19
Number of rollseg buffer pools

MULTI_PAGE_GET_NUM
1
缓冲区最多一次读取的页面数。有效值范围(1~128)

PRELOAD_SCAN_NUM
4
The number of pages scanned continuously to start preload task

PRELOAD_EXTENT_NUM
50
The number of clusters preloaded for the first time

SORT_BUF_SIZE
50
排序缓存区大小以M为单位。有效值范围(1~1024)

SORT_BLK_SIZE 1 新排序机制下,控制每次小数据量排序的内存大小

SORT_BUF_GLOBAL_SIZE 2000 新排序机制下,控制所有排序使用的内存上限

SORT_FLAG 0 值为0/1表示用老/新的方式。

HAGR_HASH_SIZE
100000
HAGR操作时,建立HASH表的桶个数。有效值范围(10000~100000000)

HJ_BUF_GLOBAL_SIZE
2000
HASH 连接操作符的数据总缓存大小(>= HJ_B…参数,以兆为单位。有效值范围(10~500000)
HJ_BUF_SIZE
250
单个HASH连接操作符的数据总缓存大小,以兆为单位。有效值范围。有效值范围(2~100000)

HJ_BLK_SIZE
1
HASH连接操作符每次分配缓存(BLK)大小,以兆为单位,必须小于HJ_BUF_SIZE。有效值范围(1~50)

HAGR_BUF_GLOBAL_SIZE
2000
HAGR、DIST、集合操作、NTTS以及HTAB操作符的数据总缓存大小(>= HAGR BUF SIZE),系统级参数以兆为单位。有效值范围(10~100000)

HAGR_BUF_SIZE
250
单个HAGR、DIST、集合操作、NTTS以及HTAB操作符的数据总缓存大小以兆为单位。有效值范围。有效值范围(2~500000)

HAGR_BLK_SIZE
1
HAGR、DIST、集合操作、NTTS以及HTAB操作符每次分配缓存(BLK)大小以兆为单位;必须小于HAGR BUF SIZE。有效值范围(1~50)

MTAB_MEM_SIZE
8
MTAB缓存BDTA占用内存空间的大小,以KB为单位(1~1048576)

FTAB_MEM_SIZE
0
FTAB缓存BDTA占用内存空间的大小,以KB为单位(O~65536)

MMT_GLOBAL_SIZE
4000
系统总共使用MMT的文件总大小,单位MB,有效值范围(10~1000000),仅在MMT_SIZE大于0时有效

MMT_SIZE
0
是否使用MMT。O:不启动;其他有效值:启用,并确定单个映射文件大小。有效值范围(O~64),单位MB

MMT_FLAG
1
MMT存储数据方式。1:按页存储;2:BDTA存储。仅在MMT_SIZE大于0时有效

DICT_BUF_SIZE
50
字典缓冲区大小,以兆为单位,有效值范围(1~2048)

HFS_CACHE_SIZE
160
HFS I/U/D 时HDTA_BUFFER缓存池大小,单位为MB。有效值范围(160~4294967294)

VM_STACK_SIZE
256
系统执行时虚拟机堆栈大小,单位为K,堆栈的空间是从操作系统中申请的,有效值范围(64~1024*1024)

VM_POOL_SIZE
256
系统执行时虚拟机内存池大小在执行过程中用到的内存大部分是从这里申请的,它的空间是从操作系统中直接申请的;有效值范围(32~1024*1024)

VM_POOL_TARGET
16384
虚拟机内存池能扩充到的最大大小,以M为单位,有效值范围(O~1010241024),0表示不限制

SESS_POOL_SIZE
16
会话缓冲区大小,以KB为单位,有效值范围(16~1024*1024)

SESS_POOL_TARGET
16384
会话缓冲区能扩充到的最大大小,以M为单位,有效值范围(0~1010241024),0表示不限制

RT_HEAP_TARGET
8192
runtime heap target size in Kilobytes

VM_MEM_HEAP
0
VM是否使用HEAP分配内存。1:是,0:否

RFIL_RECV_BUF_SIZE
16
控制服务器启动时,进行REDO操作过程中,REDO日志文件恢复时BUFFER的大小,以MB为单位,有效值范围(16~4000)

N_MEM_POOLS
1
内存池的数量,有效值范围(1~128)

COLDATA_POOL_SIZE
0
BDTA使用的共享内存池的大小,以兆为单位,有效值范围(O~100),0表示不启用COLDATA_POOL

HAGR_DISTINCT_HASH_TABLE_SIZE
10000
分组DISTINCT操作中HASH表的大小(桶数)。取值范围为(10000~100000000)

CNNTB_HASH_TABLE_SIZE
100
指定CNNTB操作符中创建HASH表的大小。值范围为(100~100000000)

GLOBAL_RTREE_BUF_SIZE
100
R树全局缓冲区大小,以MB为单位

SINGLE_RTREE_BUF_SIZE
10
单个R树的缓冲区大小,以MB为单位

SORT_OPT_SIZE
0
once max memory size of radix sort assist count array

DFS_BUF_FLUSH_OPT
0
Whether to flush buffer page in opt mode for DFS storage

线程相关参数

WORKER_THREADS
12
工作线程的数目,有效值范围(1~64)

TASK_THREADS
4
任务线程个数(1~1000)

UTHR_FLAG
0
用户线程标记,1:启用;0:不启用。

FAST_RW_LOCK
1
快速读写锁标记,1表示启用,0不启用

SPIN_TIME
4000
线程在不能进入临界区时,自旋的次数,有效值范围(O~4000)

WORK_THRD_STACK_SIZE
8192
工作线程堆栈大小,以KB为单位。有效值范围(64~4096)

WORK_THRD_RESERVE_SIZE
100
Worker Thread Reserve Stack Size In Kilobytes

WORKER_CPU_PERCENT
0
工作线程占CPU的比重,仅非WINDOWS下有效(O~100)

NESTED_C_STYLE_COMMENT
0
flag for C stype nested comment

查询先关参数

USE_PLN_POOL
1
是否重用执行计划。0:禁止执行计划的重用;1:启用执行计划的重用功能;2:进行常量参数化优化

DYN_SQL_CAN_CACHE
1
是否缓存动态语句的执行计划。0:不缓存;1:当USE_PLN_POOL不为0时,缓存动态语句的执行计划

VPD_CAN_CACHE
0
VPD SQL cache mode. 0: Forbidden; 1: Allowed if the USE_PLN_POOL is non-zero;

RS_CAN_CACHE
0
是否启用结果集缓存,0:不启用;1:启用

RS_CACHE_TABLES

指定可以缓存结果集的基表的清单,当RS_CAN_CACHE=1时,只有查询涉及的所有基表全部在此参数指定范围内;该查询才会缓存结果集

RS_CACHE_MIN_TIME
0
结果集缓存的语句执行时间下限,只有实际执行时间不少于指定时间值的查询,其结…认值0,表示不限制;有效值范围(0~4294967296),以MS为单位

RS_BDTA_FLAG
0
是否以BDTA形式返回结果集。0:以行为单位返回结果集;2:以BDTA形式返回结果集

RS_BDTA_BUF_SIZE
32
配置消息长度,单位为K。有效值范围(8~32768)

RESULT_SET_LIMIT
10000
—次请求可以生成的结果集最大个数。有效值范围(1~65000)

RESULT_SET_FOR_QUERY
0
是否不生成非查询结果集。0:生成;1:不生成

SESSION_RESULT_SET_LIMIT
10000
会话上结果集个数上限,0代表不限制(0~65000)

BUILD_FORWARD_RS
0
游标是否生成结果集。0代表生成;1代表不生成

MAX_OPT_N_TABLES
6
能优化的最大表连接个数。有效值范围(3~8)

MAX_N_GRP_PUSH_DOWN
5
Maximum Number Of Rels For Group push down Optimization

CNNTB_MAX_LEVEL
20000
层次查询的最大支持层次。有效范围(1~100000)

CTE_MAXRECURSION
100
Maximum recursive Level Of Common Expression Table

CTE_OPT_FLAG
1
Optimize recursive with, 0: false, 1: convert refed subquery to invocation

BATCH_PARAM_OPT
0
是否启用批量参数优化,0:禁用;1:启用,默认启用。当置为1时,不返回操作影响的行数

CLT_CONST_TO_PARAM
0
是否进行语句的常量参数化优化,0:不进行;1:进行

LIKE_OPT_FLAG
15
LIKE查询的优化开关,是否转换为POSITION处理。1:转换;0:不转换。取值范围:0,1

FILTER_PUSH_DOWN
0
是杏下放过滤条件到基表

USE_MCLCT
2
MCLCT USE FLAG FOR REPLACE MGAT

PHF_NTTS_OPT
1
MPP系统中是否进行NTTS计划的优化;打开时可能减少计划中的NTTS操作符。1:支持;0:不支持。取值范围:0、1

MPP_MOTION_SYNC
200
通讯操作符同步时认定的邮件堆积数,堆积超过该值则要进行同步检查。取值0-100000,0表示不进行同步检查

UPD_DEL_OPT
2
删除更新计划优化方式,取值范围0、1、2单节点计划0∶不优化;1,2…:优化删除更新计划不优化NTTS;2∶优化删除更新计划同时优化NTTS

ENABLE_INJECT_HINT
0
enable inject hint

FETCH_PACKAGE_SIZE
512
command fetch package size

UPD_QRY_LOCK_MODE
0
lock mode of FOR UPDATE query

ENABLE_DIST_IN_SUBQUERY_OPT
0
优化子查询里面的IN,O:不优化;1:优化

MAX_OPT_N_OR_BEXPS
7
能参与优化的最大OR分支个数,超过时布尔表达式仅用于过滤使用。有效值范围(7~16)

USE_HAGR_FLAG
0
当带有DISTINCT的集函数不能使用SAGR操作符时,是否使用HAGR,1:使用;0:不使用

DTABLE_PULLUP_FLAG
1
是否在语法分析阶段对派生表进行上拉优化处理,0:不优化;1:优化

VIEW_PULLUP_FLAG
是否对视图进行上拉优化,把视图转换为其原始定义,消除视图。0:不启用;1:启用

GROUP_OPT_FLAG
4
the flag of opt group

FROM_OPT_FLAG
0
the flag of opt from

HAGR_PARALLEL_OPT_FLAG
0
MPP、并行下对GROUP BY的优化开关。1:优化;0:不优化。取值范围:0、1

HAGR_DISTINCT_OPT_FLAG
2
MPP下是否对HAGR+DISTINCT进行优化,0:不优化;1:优化

REFED_EXISTS_OPT_FLAG
值范围为O或1,即TURE或FALSE。是否对相关EXISTS优化为非相关IN查询:1表示优化打开,0表示优化关闭

REFED_OPS_SUBQUERY_OPT_FLAG
0
是否将OPALL/SOME/ANY相关子查询转换为EXISTS相关子查询。0:否;1:是

HASH_PLL_OPT_FLAG
是否裁剪HASH SEMI/INNER连接右边的分区表。取值0或1。1裁剪,0不裁剪

PARTIAL_JOIN_EVALUATION_FLAG
是否对去除重复值操作的下层连接进行转换优化,0:不优化;1:优化

USE_FK_REMOVE_TABLES_FLAG
是否利用外键约束消除冗余表。0:不启用;1:启用

USE_FJ_REMOVE_TABLE_FLAG
0
Whether to remove redundant join by taking advantage of filter joining

SLCT_ERR_PROCESS_FLAG
0
How to handle error when processing single row

MPP_HASH_LR_RATE
10
MPP下,对HASH JOIN节点,可以根据左右儿子CARD代价的比值,调整HA…值超过此值,则小数据量的一方全部收集到主EP来做。取值范围(10~10000)

LPQ_HASH_LR_RATE
30
LPQ下,对HASH JOIN节点,可以根据左右儿子CARD代价的比值,调整HASH_JOIN的左右儿子的MOTIOR添加,从而影响计划。

USE_HTAB
1
Whether to use HTAB operator for the whole plan

SEL_ITEM_HTAB_FLAG
0
当查询项中有相关子查询时,是否做HTAB优化。0:不优化;1:优化。

OR_CVT_HTAB_FLAG
1
当查询条件OR中含有公共因子时,是否允许使用HTAB来进行优化。1:使用;0:不使用。取值范围:0、1

ENHANCED_SUBQ_MERGING
0
Whether to use merging subquery opt

CASE_WHEN_CVT_IFUN
9
是否将CASE WHEN THEN ELSE END语句转换为IFOPERATOR函数。0:不转换:1∶转换;2∶转换,且有限制地进行表达式重用

OR_NBEXP_CVT_CASE_WHEN_FLAG
0
是否将OR转化为CASE WHEN THEN ELSE END语句。0:不转换;1:转换

NONCONST_OR_CVT_IN_LST_FLAG
0
是否开启IN LIST优化:将不含有常量的OR表达式转换成IN LIST。0:不开启;1:开启

OUTER_CVT_INNER…L_UP_COND_FLAG
1
当外连接转化为内连接时,是否打开连接条件。0:不打开;1:打开

OPT_OR_FOR_HUGE_TABLE_FLAG
0
是否使用HFSEK优化HUGE表中列的OR过滤条件。1:使用;0:不使用。取值范围:0、1

ORDER_BY_NULLS_FLAG
0
ASC升序排序时,控制NULL值返回的位置。取值0或1。1表示NULL值在…情况下,NULL值的返回与ORACLE保持一致。DESC降序时该参数无效

SUBQ_CVT_SPL_FLAG
1
控制相关子查询的实现方式

ENABLE_RQ_TO_SPL
1
是否将相关子查询转换为SPOOL方式,0:不优化;1:优化

MULTI_IN_CVT_EXISTS
0
多列IN是否转换为等价的EXISTS过滤,取值范围:0、1

PRJT_REPLACE_NPAR
1
是否将引用列转换为变量VAR,并替换查询表达式中的引用列,0:不优化;1:优化

ENABLE_RQ_TO_INV
0
相关查询表达式是否转换为函数方式实现,取值范围:0、1

SUBQ_EXP_CVT_FLAG
0
WHETHER CONVERT REFERED SUBQUERY EXP TO NON-REFERED SUBQUERY EXP

USE_REFER_TAB_ONLY
0
处理相关子查询时是仅将相关的表下放,或者连同上方的SEMI/HASH JOIN一起下放。0:一起下放;1:仅下放相关表

REFED_SUBQ_CROSS_FLAG
1
是否将相关子查询与外层表优化为CROSS JOIN,0:不优化;1:优化

IN_LIST_AS_JOIN_KEY
0
搜索多表连接方式时,对于索引连接(INDEX JOIN)的探测,NEXP_IN_IST表达式类型可以作为多表连接的KEY。

OUTER_JOIN_FLATING_FLAG
0
优化外连接使得类似这样的连接ALET(8 CROSS C) ON AC1=B.C1优化为ALEFT(8.1 CROSS C ON A.C1=B.C1,相当于把A和日的过族器平坦化到了下层查询中使得8获取的中间结果较小

TOP_ORDER_OPT_FLAG
1
优北带有TOP和ORDER BY子句的查询使得SORT操作符可以省路。优化的效果是尽量使得OR …TM作符减少排序操作。如果排序列不届于同一个基表或者排序列不是基表列则肯定是不可以优化

PLACE_GROUP_BY_FLAG
0
在含有集函数的查询中,是否允许先分组减少数据量后再进行连接。0:不允许;1:允许

TOP_DIS_HASH_FLAG
0
是否通过禁用HASHJOIN方式来优化TOP查询取值: 0:不优化1:当OPTIMZER MODE分0时, …禁用HASHJION兰OPTIMER MODE为1时TOP下方最近的连接倾向于不使用HASH ION

ENABLE_RQ_TO_NONREF_SPL
0
相关查询表达式转化为非相关查询表达式,目的在于相关查询表达式的执行处理由之前的平坦化方式转化为一行一行处理.类似ORACLE的每行处理机制。

OPTIMIZER_MODE
1
DM优化器的模式,0:老优化器模式;1:新优化器模式

NEW_MOTION
0
NEW_MOTION

LDIS_NEW_FOLD_FUN
0
ldis use different fold fun with mdis

DYNAMIC_CALC_NODES
0
different nodes of npln use different nubmer of calc sizes/threads

OPTIMIZER_MAX_PERM
7200
控制计划探测过程中的最大排列数,如果大于该阙值,则减少相应的排列,减少探测计划的时间,有效值范围(1~4294967293)

ENABLE_INDEX_FILTER
0
是否进行索引过滤优化,可取值为0、1、2

OPTIMIZER_DYNAMIC_SAMPLING
0
当统计信息不可用时是否启用动态统计信息。取值范围:012,0:不启用110:启用采用率10%100%,11:启用由优化器确定采样率(0.1%99.8%)12:同11,但收集的结果会持久化保存

TABLE_STAT_FLAG
0
How to use stat of table

AUTO_STAT_OBJ
0
Flag of automatically collecting statistics and recording DML changing rows

NONREFED_SUBQUERY_AS_CONST
0
是否将非相关子查询转化为常量处理。O:不进行优化,对非相关子查询使用连接方式处理:1:将非相关子查询转换为常量,作为过滤条件使用

HASH_CMP_OPT_FLAG
0
是否启用静态哈希表的优化有效取值0124.0不启用优化。1对HAGR分组计算中带有CS…只需把对应的数值相加。例如需要启用HAGR分组计算和HASH连接的相关优化如则可以把该参数设置为6。

OUTER_OPT_NLO_FLAG
0
外连接或内连接是否支持右孩子为非基表情况下的NDEXION O0:不启用;1:启用外连接下的优化…用内连接下的优化;3:同时启用内连接和外连接。该参数仅在OPIMZER、 MODE =0时生效

DISTINCT_USE_INDEX_SKIP
2
DSTNCT例是否使用索引的跣扫辙(单列]索引或复台索引)。专门用于SQ语句中有DSTINCT的.场…0:不使用1:强制使用2∶根据代价选法择是否使用。该参数只在OPTIMZER MODE=1时有效

USE_INDEX_SKIP_SCAN
0
I间隔索引扫描的开关

INDEX_SKIP_SCAN_RATE
0.003
间隔索引扫描的代价调节开关。当列的 DISTINCT数/总行数大于该值时,就不再使用间隔索引扫描的方式

SPEED_SEMI_JOIN_PLAN
1
是否加速半连接的探测过程。0:不加速;1:加速计划探测;2:加速计划探测和执行

COMPLEX_VIEW_MERGING
0
对于复杂视图会执行合并操作使得GROUP分组操作在连接之后才执行。O:不启用1:对不包含别名和同名列的视图进行合并;2∶视图定义包含别名或同名列时也进行合并

HLSM_FLAG
1
控制多列非相关NOT IN的查询实现方式

DEL_HP_OPT_FLAG
0
控制分区表的操作优化

OPTIMIZER_OR_NBEXP
0
OR表达式的优化方式

CNNTB_OPT_FLAG
0
是否使用优化的层次查询执行机制。0:不使用;1:强制使用;2:优化器自动决定是否使用

ADAPTIVE_NPLN_FLAG
0
是否启用自适应计划机制仅OPITIMIZER MOOE=1时生效。0:不启用)1:对索引L连接、嵌套含. .生接自用自适应计划2∶ORDER BV在HASH连接时启用自适应计划3:同时启用1和2的优化机制

MULTI_UPD_OPT_FLAG
0
是否使用优化的多列更新。0:不使用,仍按照语句改写方式实现;1:利用多列SPL功能加以实现

MULTI_UPD_MAX_COL_NUM
128
Max value of column counts when optimize multi column update

ENHANCE_BIND_PEEKING
0
是否使用自适应的绑定变量窥探开关。0:不适用;1:使用

NBEXP_OPT_FLAG
3
Whether to enable optimization for bool expressions

HAGR_HASH_ALGORITHM_FLAG
0
HAGR中HASH算法选择标记。0:按照移位方式计算;:1:按照异或方式计算

DIST_HASH_ALGORITHM_FLAG
0
DISTINCT中HASH算法选择标记。0:按照移位方式计算;1:按照异或方式计算

UNPIVOT_ORDER_FLAG
0
是否对UNPIVOT的结果按照公共列排序

VIEW_FILTER_MERGING
10
是否对视图条件进行合并优化以及如何优化,0:不优化;1:尽可能地进行视图条件合并;2:自动判断是否进行视图条件合并

ENABLE_PARTITION_WISE_OPT
0
是否利用水平分区表的分区信息进行排序、分组的计划优化,即在满足条件的情况下对每个子表单独排序、分组。0:不启用;1:启用

OPT_MEM_CHECK
0
内存紧张时,优化器是否缩减计划探测空间,仅OPTIMIZER_MODE=0时生效。0:不缩减计划探测空间;1:缩减计划探测空间

ENABLE_JOIN_FACTORIZATION
0
是否启用连接分解,即UNION ALL分支间存在公共部分时是否进行公因子提取。0:不启用;1:启用

EXPLAIN_SHOW_FACTOR
1
执行计划显示的范围因子。取值范围(1~100)

ERROR_COMPATIBLE_FLAG
0
是否对子查询同名列进行报错。取值O或1。0报错,1不报错

ENABLE_NEST_LOOP_JOIN_CACHE
0
是否考虑缓存嵌套循环连接中间结果以加速执行。取值O或1。0:不启用;1:启用

检查点相关参数

CKPT_RLOG_SIZE
100
产生多大日志文件后做检查点,以兆为单位。有效值范围(0~4294967294)

CKPT_DIRTY_PAGES
10000
产生多少脏页后产生检查点,以页为单位。有效值范围(O~4294967294)

CKPT_INTERVAL
300
指定检查点的时间间隔。以秒为单位,为0时表示不自动定时做检查点。有效值范围(0~2147483647)

CKPT_FLUSH_RATE
5.00
检查点刷盘比例。有效值范围(1~100)

CKPT_FLUSH_PAGES
1000
检查点刷盘的最小页数。有效值范围(1000~100000)

CKPT_WAIT_PAGES
128
检查点一次发起的最大写入页数,有效值范围(1~1024)

FORCE_FLUSH_PAGES
8
调度线程启动刷脏页流程时,每个BUFFER POOL写入磁盘的脏页数。有效值范围(O~1000)

IO相关参数

DIRECT_IO
0
非WINDOWS下有效,0:使用文件系统缓存;1,2:不使用文件系统缓存

IO_THR_GROUPS
4
非WINDOWS下有效,表示IO线程组个数。有效值范围(O~512)

HIO_THR_GROUPS
2
HUGE缓冲区IO线程组数目。有效值范围(1 ~512)

FAST_EXTEND_WITH_DS
0
How To Extend File’s Size (Non-Windows Only), 0: Extend File With Hole; 1: Extend File With Disk Space

数据库相关参数

MAX_SESSIONS
1500
系统允许同时连接的最大数,同时还受到LICENSE的限制,取二者中较小的值,有效值范围(1~20000)

MAX_CONCURRENT_TRX
0
表示系统支持同时运行事务数的最大值。有效值范围(O~150O),0表示不控制

CONCURRENT_TRX_MODE
0
Concurrent transactions mode

TRX_VIEW_SIZE
512
The buffer size of local transaction ids in TRX_VIEW

MAX_SESSION_STATEMENT
20000
单个会话上允许同时打开的语句句柄最大数,有效值范围(64~20480)

MAX_CONCURRENT_OLAP_QUERY
0
OLAP模式下能同时执行的复杂查询个数,超过限制后;复杂查询将在执行阶段进入等待。0表示不做限制。有效值范围(0~100)

BIG_TABLE_THRESHHOLD
1000
OLAP环境下对复杂查询最大并发数(参数MAXCONCURRENT_OLAP _QUERVY)进行限制时认定是否为复杂查询中表的行数下限以万为单位。取值范围(0~4294967294)

MAX_EP_SITES
64
MPP环境下EP站点的最大数量,有效值范围(2~1024)

PORT_NUM
5236
服务器通讯端口号,有效值范围(1024~65534)

LISTEN_IP

IP address from which the database server can accept

FAST_LOGIN
0
是否在登录时记录登录历史信息。0记录,1不记录

DDL_AUTO_COMMIT
1
指定DDL语句是否自动提交,1:自动;0:手动。当COMPATIBLE_MODE=1时,DDL_AUTO_COMMIT的实际值均为0。

COMPRESS_MODE
0
建表时是否缺省压缩。O:不进行;1:进行

PK_WITH_CLUSTER
0
建立主关键字时,是否缺省指定为CLUSTER,0:不指定;1:指定

EXPR_N_LEVEL
200
表达式最大嵌套层数。有效值范围(10~1000)

N_PARSE_LEVEL
100
表示对象PROC、VIEW、PKG、CLASS的最大解析层次,如果层次过深则报错返回。有效值范围(30~1000)

MAX_SQL_LEVEL
500
Maximum nesting levels of VM stack frame for sql

BDTA_SIZE
1000
BDTA缓存的记录数。有效值范围(1~10000)

OLAP_FLAG
2
启用联机分析处理,0:不启用;1:启用

JOIN_HASH_SIZE
500000
HASH JOIN操作时,HASH表大小,以CELL个数为单位。有效值范围(O~100000000)

HFILES_OPENED
256
设置可以同时打开的列存储表数据文件个数。有效值范围(60-10000)

ISO_IGNORE
0
是否忽略显示设置的事务隔离级别,1表示忽略,0不忽略

TEMP_SIZE
1024
临时数据库大小,以兆为单位。有效值范围(10~1048576)

TEMP_SPACE_LIMIT
102400
临时表空间大小上限,以M为单位。0表示不限制临时表空间大小。有效范围(O~ 4294967294)

FILE_TRACE
0
日志中是否记录文件操作,0:不记录;1:记录

COMM_TRACE
1
服务器日志是否记录通信中产生的警告信息。0:不记录;1:记录

ERROR_TRACE
0
Whether to log error information, 1: NPAR ERROR

CACHE_POOL_SIZE
1024
SQL缓冲池大小,以兆为单位。有效值范围(1~10240)

PLN_DICT_HASH_THRESHOLD
20
Threshold in megabytes for plan dictionary hash table creating

STAT_COLLECT_SIZE
10000
统计信息收集时,样本的最小行数。有效值范围(O~10000000)

STAT_ALL
0
在估算分区表行数时,是否采样所有分区子表。0:不采样所有;1:采样所有

PHC_MODE_ENFORCE
0
允许使用的JOIN实现方式,0:所有;1:允许NEST LOOP;2:允许INDEX JOIN;4:允许HASH JOIN;8:允许MERGE JOIN
ENABLE_HASH_JOIN
0
是否允许使用哈希连接,0:不允许;1:允许。

ENABLE_INDEX_JOIN
1
是否允许使用索引连接,0:不允许;1:允许。

ENABLE_MERGE_JOIN
1
是否允许使用归并连接,0:不允许;1:允许

MPP_INDEX_JOIN_OPT_FLAG
0
MPP下,是否可以使用索引左外连接,0:不使用;1:使用

MPP_NLI_OPT_FLAG
0
MPP下嵌套注接的优化方式(0:嵌套连接操作的左右孩子添加MPGATHER.汇总到主EP执行,父节点添加MPSCATTER3:嵌赛连接操作的左右孩子都添加MPPGATHER和MPP SCATTR

MAX_PARALLEL_DEGREE
1
用来设置最大并行任务个数。取值范围:1~128。缺省值1,表示无并行任务。当PARALLEL_POLICV值为1或2时该参数值才有效。

PARALLEL_POLICY
0
用来设置并行策略。取值范围0、1和2,缺省为0。其中,0表示不支持并行;1表示自动配置并行工作线程个数(与物理CPU核数相同);2表示手动设置并行工作线程数

PARALLEL_THRD_NUM
10
用来设置并行工作线程个数,仅当PARALLEL_POLICY值为2时才启用此参数。有效值范围(1~1024)

PARALLEL_MODE_COMMON_DEGREE
1
并发模式程度。此值越大,扫描代价将越大,越倾向于使用索引连接而不是哈希连接,需酌情使用。

PUSH_SUBO
0
是否下放子查询使先做。O:相关子查询不下放非相关子查询在不存在单表过滤条件时下放;1:不存在单表过浦条件时下放;2:始终考虑下放

OPTIMIZER_AGGR_GROUPBY_ELIM
1
优化器被允许的连接方式。1:嵌套循环连接;2:索引连接;4:哈希连接;8:归并连接还可使用以上值的组合值

UPD_TAB_INFO
0
系统启动时是否更新表信息(如行数),1:启用更新;O:不更新。

ENABLE_IN_VALUE_LIST_OPT
6
是否允许INIST表达式优化O不优化:1:将HINLIST表达式在语义分析阶段优化为CONSTV处理; 2在查询计划确定后把INLST表达式转化为CONSTV3洞时进行置为1和2时的优化处理

ENHANCED_BEXP_TRANS_GEN
3
是否允许非等值布尔表达式和外连接ON条件中的传递闭包。0:不允许;1:允许

ENABLE_DIST_VIEW_UPDATE
0
是否支持更新含有DISTINCT的视图,1:允许;0:不允许,当COMPATIBLE_MODE=1时,ENABLE_DIST_VIEW_UPDATE的实际值均为1

STAR_TRANSFORMATION_ENABLED
0
是否允许对星形模型查询改写以使用位图连接索引。0:不启用;1:启用

MONITOR_INDEX_FLAG
0
是否对索引进行监控O:关闭自动监控,可使用ALTER INDEX语句启用索引监控;1∶打开自动监控.对用户定义的二级索引进行监控;2∶禁止索引监控

AUTO_COMPILE_FLAG
1
Whether to compile the invalid objects when loading

RAISE_CASE_NOT_FOUND
O
Whether raise CASE_NOT_FOUND exception for no case item matched

FIRST_ROWS
100
结果集一个消息中返回的最大行数(1~1000)

LIST_TABLE
0
默认情况下,创建的表是否为LIST表,0:否;1:是

ENABLE_SPACELIMIT_CHECK
1
是否启用检查SPACE LIMIT。1:启动;0:不启用

BUILD_VERTICAL_PK_BTREE
0
创建VERTICAL表的时候,是否创建主键。1:创建,0:不创建

BDTA_PACKAGE_COMPRESS
0
是否启动BDTA压缩传递功能,1:压缩;0:不压缩

HFINS_PARALLEL_FLAG
0
是否启动HUGE表查询插入优化,1:是;0:否

HFINS_MAX_THRD_NUM
100
HUGE表查询插入优化时并行的最大线程数,有效值范围(4~200)

LINK_CONN_KEEP_TIME
15
DBLINK空闲连接保持时间,以分钟为单位,为0时表示不主动释放空闲DBLINK连接

DETERMIN_CACHE_SIZE
5
确定性函数值缓冲区大小,以MB为单位

NTYPE_MAX_OBJ_NUM
1000000
复合数据类型中包含的对象或者字符串的总个数,以及查询中包含的变量的总个数,有效值范围为(10,100000000)

CTAB_SEL_WITH_CONS
0
查询建表时,是否对原始表上的约束进行拷贝。1:是,0:否

HLDR_BUF_SIZE
8
进行HUGE表导入时,缓冲区的大小单位MB,有效值范围(4~1024)

HLDR_BUF_TOTAL_SIZE
4294967294
HUGE table fast loader buffer total size in Megabytes

HLDR_REPAIR_FLAG
0
Flag of repairing huge table after exception, 0: NO 1: YES

HLDR_FORCE_COLUMN_STORAGE
1
Whether force column storage for last section data, 0: NO 1: YES

HLDR_FORCE_COL…ORAGE_PERCENT
80
HLDR_FORCE_COLUMN_STORAGE_PERCENT

HLDR_HOLD_RATE
1.50
HLDR_HOLD_RATE

HLDR_MAX_RATE
2
HLDR_MAX_RATE

HUGE_ACID
0
是否支持HUGE表增删改与查询的并发,取值范围:0、1、2

HUGE_STAT_MODE
2
HUGE_STAT_MODE

HFS_CHECK_SUM
1
HFS_CHECK_SUM

HBUF_DATA_MODE
0
HUGE表数据缓冲区中缓存数据的格式,取值范围:0、1

DBLINK_OPT_FLAG
509
DBLINK_OPT_FLAG

FILL_COL_DESC_FLAG
0
服务器返回给客户端的结果集是否含有列描述信息,取值范围:0、1,默认值为1

BTR_SPLIT_MODE
0
B树叶子节点分裂方式

BLOB_OUTROW_REC_STOR
5
大数据行外记录式存储配置参数

TS_RESERVED_EXTENTS
64
系统为每个表空间提前预留的簇个数,减少系统在执行过程中申请不到空间的情况。有效值范围(2~1024)

TS_SAFE_FREE_EXTENTS
512
系统认为安全的FREE EXTENTS空间,有效值范围(128~65534)

TS_MAX_ID
8192
限制系统所支持的最大表空间ID,有效值范围(5~65517),与系统实际最大ID比较,取大

TS_FIL_MAX_ID
255
限制每个表空间所支持的最大文件个数,范围(2~255)

DECIMAL_FIX_STORAGE
0
DECIMAL_FIX_STORAGE

SQL_SAFE_UPDATE_ROWS
0
SQL_SAFE_UPDATE_ROWS

ENABLE_HUGE_SECIND
1
ENABLE_HUGE_SECIND

TRXID_UNCOVERED
0
TRXID_UNCOVERED

LOB_MAX_INROW_LEN
900
LOB_MAX_INROW_LEN

RS_PRE_FETCH
0
RS_PRE_FETCH

DFS_PAGE_FLUSH_VALIDATE
1
DFS_PAGE_FLUSH_VALIDATE

GEN_SQL_MEM_RECLAIM
1
GEN_SQL_MEM_RECLAIM

TIMER_TRIG_CHECK_INTERVAL
60
TIMER_TRIG_CHECK_INTERVAL

DFS_HUGE_SUPPORT
1
DFS_HUGE_SUPPORT

INNER_INDEX_DDL_SHOW
1
INNER_INDEX_DDL_SHOW

PWR_FLUSH_PAGES
1000
控制生成特殊PWR记录的频率,每写入PWR FLUSH PAGES页到磁盘,即生成一条特殊的PWR记录,标记之前的PWR记录是有效的。有效值范围(0~4294967294)

IGNORE_FILE_SYS_CHECK
1
系统重启时,是否检查SVSTEM/ROLL/MAIN表空间文件系统;取值范围0,1;默认O:检查文件系统,1:不检查文件系统

FILE_SCAN_PERCENT
100.00
FILE_SCAN_PERCENT

STARTUP_CHECKPOINT
0
系统启动时REDO后是香强制做完全检查点。0:不强制;1:强制

CHECK_SVR_VERSION
1
数据库记录的执行码版本比当前SERVER版本高时,是否报错。0:不报错;1:报错,服务器不能启动成功

ID_RECYCLE_FLAG
0
ID_RECYCLE_FLAG

BAK_USE_AP
1
备份还原是否使用DMAP插件

ENABLE_DBLINK_TO_INV
0
ENABLE_DBLINK_TO_INV

ENABLE_BLOB_CMP_FLAG
0
ENABLE_BLOB_CMP_FLAG

ENABLE_CREATE_BM_INDEX_FLAG
1
ENABLE_CREATE_BM_INDEX_FLAG

CVIEW_STAR_WITH_PREFIX
1
CVIEW_STAR_WITH_PREFIX

ENABLE_ADJUST_NLI_COST
1
ENABLE_ADJUST_NLI_COST

ENABLE_SEQ_REUSE
0
ENABLE_SEQ_REUSE

REDO日志相关参数

ELOG_REPORT_LINK_SQL
0
是否记录DBLINK执行的SQL到服务器日志文件中。0:不记录;1:记录

RLOG_BUF_SIZE
1024
单个日志缓冲区大小(以日志页个数为单位),取值只能为2的次幂值,最小值为1,最大值为20480

RLOG_POOL_SIZE
1024
最大日志缓冲区大小(以M为单位)。有效值范围(16~1024)

RLOG_PARALLEL_ENABLE
0
是否启动并行日志,1:启用;0:不启用

RLOG_IGNORE_TABLE_SET
1
RLOG_IGNORE_TABLE_SET

RLOG_APPEND_LOGIC
1
是否启用在日志中记录逻辑操作的功能。1:启用;0:不启用

RLOG_APPEND_SYSTAB_LOGIC
0
是否启用在日志中记录系统表逻辑操作的功能,启用RLOG_APPEND_LOGIC后有效,取值范围0、1,0:不启用;1:启用

RLOG_SAFE_SPACE
128
安全的可用日志空间大小(以M为单位)。有效值范围(0 ~ 1024)

RLOG_SEND_APPLY_MON
64
数据守护中对于主库用于指定统计最近N次主库到每个备库的归档发送时间;对于备库,用于指定统计最近N次备库重演日志的时间N为此参数设置的值。有效值范围(16~1024)

REDO_PWR_OPT
1
系统故障重启时,是否启动PWR优化;取值范围0,1

REDO_IGNORE_DB_VERSION
0
启动重做REDO日志时,是否检查版本信息

REDO_BUF_SIZE
64
REDO_BUF_SIZE

REDOS_BUF_SIZE
1024
备库日志堆积的内存限制堆积的日志BUF占用内存超过此限制则延迟响应等待重演释放部分内存后再响应。以兆为单位有效值范围(0~65536M),0表示无内存限制

REDOS_MAX_DELAY
1800
备库重演日志BU/的时印段制超过此痰制则认为重演异常服务器自动宕机防止日志堆积、主库不能及时响应用户请求。以秒为单位.取值范围(0~7200)。0表示无重做时间限制

REDOS_BUF_NUM
4096
备库日志BUF允许堆积的数目限制,超过限制则延迟响应主库,等待堆积数减少后再响应。以个数为单位.有效值范围(0~9999)

REDOS_PRE_LOAD
32
备库重演日志时预加载的日志BUF个数。取值范围(0~4096)

REDOS_PARALLEL_NUM
1
REDOS_PARALLEL_NUM

预先装载表相关参数

LOAD_TABLE

在服务器启动时预先装载的表的完整表名,即*模式名.表名",多个表之间用逗号分隔,最多可指定10个表

LOAD_HTABLE

在服务器启动时预先装载的HUGE表的完整表名,即"模式名.表名",多个表之间用逗号分隔

CLT_CACHE_TABLES

指定可以在客户W缓存的表。表名必须带模式名前缀如果表名或棋式名中包会特殊字符,需要使用双引号包 .数值太长导致N文件分析困难允许在IN文件中设置多行CLT_CACHE TABLES参数

日志相关参数

LOG_POOL_SIZE
1024
最大日志缓冲区大小(以M为单位)。有效值范围(16~1024)

RLOG_COMPRESS_LEVEL
0
RLOG_COMPRESS_LEVEL

RLOG_ENC_CMPR_THREAD
4
RLOG_ENC_CMPR_THREAD

RLOG_PKG_SEND_ECPR_ONLY
0
RLOG_PKG_SEND_ECPR_ONLY

DFS_RLOG_SEND_POLICY
1
DFS_RLOG_SEND_POLICY

事务相关参数

ISOLATION_LEVEL
1
系统默认隔离级别。1:读提交;3:可串行化

DDL_WAIT_TIME
10
DDL操作的锁超时时间,以秒为单位。有效值范围(O~60)

BLDR_WAIT_TIME
10
BLDR_WAIT_TIME

MPP_WAIT_TIME
10
MPP_WAIT_TIME

FAST_RELEASE_SLOCK
1
是否启用快速释放S锁,1:启用;0:不启用

SESS_CHECK_INTERVAL
3
循环检测会话状态的时间间隔,以秒为单位。有效值范围(1~60)

LOCK_TID_MODE
1
SELECT FOR UPDATE封锁方式。0:结果集记录小于100行,直接封锁TID,否则升级为表锁;1:不升级表锁,—律使用TID锁

LOCK_TID_UPGRADE
0
LOCK_TID_UPGRADE

NOWAIT_WHEN_UNIQUE_CONFLICT
0
插入数据时,如果和未提交数据有UNIQUE约束的冲突,是否等待未提交事务结束,0:等待,直至未提交事务结束;1:不等待,立即返回错误

UNDO_EXTENT_NUM
4
表示系统启动时,为每个工作线程分配的回滚簇个数。有效值范围(1~256)

MAX_DE_TIMEOUT
10
C外部函数的执行超时时间,以秒为单位。有效值范围(1~3600)

TRX_RLOG_WAIT_MODE
0
TRX_RLOG_WAIT_MODE

TRANSACTIONS
75
指定一个会话中可以并发的自治事务数量。取值范围(1~1000)

MVCC_RETRY_TIMES
5
指定发生MVCC冲突时的最大重试次数。有效值范围(1~4294967294)

MVCC_PAGE_OPT
1
MVCC_PAGE_OPT

ENABLE_FLASHBACK
0
是否启用闪回查询,0:不启用;1:启用

UNDO_RETENTION
90.000
事务提交后回滚页保持时间,单位为秒。有效值范围(1~86400)

PARALLEL_PURGE_FLAG
0
PARALLEL_PURGE_FLAG

PURGE_WAIT_TIME
500
检测到系统清理动作滞后(待清理事务提交时间-当前系统时间>UNDO_RETENTION)情况下,系统等待时间(MS)。有效范围(O~~60000),0表示不等待

PSEG_RECV
3
系统故障重启时,是否回滚活动事务并PURGE已经提交事务。1:是;0:否
0:跳过回滚活动事务和 PURGE 已经提交事务的步骤。
1:回滚活动事务并PURGE 已经提交事务;
2:延迟 PURGE 已提交事务,延迟回滚活动事务;
3:回滚活动事务,延迟 PURGE 已提交事务。

ENABLE_IGNORE_PURGE_REC
0
当返回EC_RN_NREC_ PURGED(-7120)错误(回滚记录版本太旧,无法获取用户记录)时的处理策略;1∶报错;O:忽略这一条记录;继续执

ENABLE_TMP_TAB_ROLLBACK
1
临时表操作是否生成回滚记录,0:不生成; 1:生成。注:置为0时,临时表的DML操作无法回滚

ROLL_ON_ERR
0
服务器执行出错时的回滚策略选择,0:回滚当前语句;1:回滚整个事务

XA_TRX_IDLE_TIME
60
XA_TRX_IDLE_TIME

LOB_MVCC
1
LOB_MVCC

LOCK_DICT_OPT
2
LOCK_DICT_OPT

TRX_DICT_DICT_NUM
64
TRX_DICT_DICT_NUM

安全相关参数

PWD_POLICY
2
设置系统默认口令策略。0:无策略1:禁止与用户名相同2口令长度不小于64:至少包含一个大写字母(A-2)84至少包含一个数字(-9)16至少包含一个抗点将号英文输入法状态下赊和空格外的所有

PWD_MIN_LEN
9
PWD_MIN_LEN

ENABLE_ENCRYPT
0
通讯所采用的加密类型。0:不加密;1:SSL加密

CLIENT_UKEY
0
CLIENT_UKEY

MIN_SSL_VERSION
771
MIN_SSL_VERSION

ENABLE_UDP
0
是否支持UDP。0:否;1:单发单收;2:多发多收。缺省使用TCP

UDP_MAX_IDLE
15
UDP最大等待时间,单位秒,超过则认为连接断开。取值范围(5~60)

UDP_BTU_COUNT
8
UDP最大不等待连续发送消息块数,当不连续差值大于它时则等待发送。取值范围(4~32)

ENABLE_IPC
0
ENABLE_IPC

AUDIT_FILE_FULL_MODE
1
审计文件满后的处理方式,1:删除文件;2:不删除文件,也不添加审计记录

AUDIT_MAX_FILE_SIZE
100
审计文件的最大大小,以兆为单位。有效值范围(1~4096)

AUDIT_IP_STYLE
0
AUDIT_IP_STYLE

ENABLE_OBJ_REUSE
0
是否支持客体重用,0:不支持;1:支持。(该参数设置仅安全版有效)

ENABLE_REMOTE_OSAUTH
0
是否支持远程操作系统认证,0:不支持;1:支持。(该参数设置仅安全版有效)

ENABLE_LOCAL_OSAUTH
0
ENABLE_LOCAL_OSAUTH

MSG_COMPRESS_TYPE
2
是否加载压缩动态库,0:不加载;1:加载

ENABLE_STRICT_CHECK
0
是否检查存储过程中EXECUTE IMMEDIATE语句的权限,1:表示检查;0:不检查

MAC_LABEL_OPTION
1
用于控制SF_LABEL_FROM_CHAR函数的使用范围。

LDAP_HOST

LDAP服务器LIP

COMM_ENCRYPT_NAME

消息加密算法名。如果为空则不进行通信加密;如果给的加密算法名错误则用默认加密算法DES_CFB。DM支持的加密算法名可以通过查询动态视图V$CIPHERS获取

COMM_VALIDATE
1
是否对消息进行校验。0不检验;1检验

MESSAGE_CHECK
0
MESSAGE_CHECK

ENABLE_EXTERNAL_CALL
0
是否允许创建或执行外部函数。0:不允许;1:允许

EXTERNAL_JFUN_PORT
6363
EXTERNAL_JFUN_PORT

EXTERNAL_AP_PORT
4236
EXTERNAL_AP_PORT

ENABLE_PL_SYNONYM
0
ENABLE_PL_SYNONYM

ENABLE_DDL_ANY_PRIV
1
ENABLE_DDL_ANY_PRIV

FORCE_CERTIFICATE_ENCRYPTION
0
FORCE_CERTIFICATE_ENCRYPTION

SEC_PRIV_MODE
0
SEC_PRIV_MODE

REGEXP_MATCH_PATTERN
0
REGEXP_MATCH_PATTERN

兼容性相关参数

BACKSLASH_ESCAPE
0
语法分析对字符串中的反斜杠是否需要进行转义处理,0:不进行转义处理;1:进行转义处理

STR_LIKE_IGNORE…ATCH_END_SPACE.1、
1
LIKE运算中是否忽略匹配串的结尾0。0:不忽略;1:忽略

CLOB_LIKE_MAX_LEN
10240
LIKE语句中CLOB类型的最大长度,单位KB,有效值范围(8~102400)

EXCLUDE_DB_NAME

服务器可以忽略的数据库名列表,各数据库名以逗号","隔开,数据库名不需要加引号.

MS_PARSE_PERMIT
O
是否支持MS SQLSERVER的语法。0:不支持;1:支持。当COMPATIBLE_MODE=3时,MS_PARSE_PERMIT的实际值为1

COMPATIBLE_MODE
0
是否兼容其他数据库模式。O:不兼容,1:兼容SQL92标准,2:兼容ORACLE,3:兼容MS SQL SERVER,4:兼容MYSQL

DATETIME_FMT_MODE
0
DATETIME_FMT_MODE

DOUBLE_MODE
0
DOUBLE_MODE

CASE_COMPATIBLE_MODE
1
涉及不同数据类型的CASE运算,是否需要兼容ORACLE的处理策略。0:不兼容;1:兼容

EXCLUDE_RESERVED_WORDS

语法解析时,需要去除的保留字列表,保留字之间以逗号分隔

COUNT_64BIT
1
COUNT集函数的值是否设置为BIGINT。0:否;1:是

CALC_AS_DECIMAL
0
0:默认值,表示按照现有逻辑处理

CMP_AS_DECIMAL
0
0:默认值,表示不对字符串与整型的比较结果类型做任何修改

CAST_VARCHAR_MODE
1
在字符串向整型转换进行比较时,是否将溢出值转为特殊伪值。0:香;1:是

PL_SQLCODE_COMPATIBLE
0
默认值为0;如果设置为1,则PL的异常处理中,SQLCODE的错误码值需要尽量与ORACLE一致

LEGACY_SEQUENCE
0
序列兼容参数,0:与ORACLE兼容;1:与旧版本DM兼容,不推荐使用

DM6_TODATE_FMT
0
TO_DATE中的HH格式小时制,0:24小时制;1:12小时制

PK_MAP_TO_DIS
0
PK_MAP_TO_DIS

PROXY_PROTOCOL_MODE
0
PROXY_PROTOCOL_MODE

DROP_CASCADE_VIEW
0
删除表或者视图的时候级联删除视图,0:只删

跟踪相关参数

SVR_LOG_NAME
SLOG_ALL
使用SQLLOG.INI中预设的模式的名称

SVR_LOG
0
是否打开SQL日志功能0表示关闭)1表示日志文件非切换模式但输出的日志格式是详细模式2表示日志文件为切换模式,输出的日志也是详细模式3表示日志为非切换模式但输出日志为简单模式

GDB_THREAD_INFO
0
系统强制HALT时,是否打印线程堆栈信息到日志文件中。0:不打印;1:打印

TRACE_PATH

存放系统TRACE文件的路径

SVR_OUTPUT
0
SVR_OUTPUT

ENABLE_ECS
0
ENABLE_ECS

ECS_HB_INTERVAL
60
ECS_HB_INTERVAL

EP_GROUP_NAME
DMGROUP
EP_GROUP_NAME

AP_GROUP_NAME
DMGROUP
AP_GROUP_NAME

DCRS_IP
LOCALHOST
DCRS_IP

DCRS_PORT_NUM
6236
DCRS_PORT_NUM

AP_IP
LOCALHOST
AP_IP

AP_PORT_NUM
0
AP_PORT_NUM

AP_PARALLEL_DEGREE
10
AP_PARALLEL_DEGREE

ENABLE_ECS_MSG_CHECK
0
ENABLE_ECS_MSG_CHECK

动态监控相关参数

ENABLE_MONITOR
1
启用动态监控功能标记,0:不启用;1:低级别监控;2:高级别监控;3:在高级别监控的基础上增加搜集各操作符的执行时间

MONITOR_TIME
1
用于打开或者关闭时间监控。该监控项的生效必须是在ENABLE_MONITOR打开的情况下。1:打开;0:

MONITOR_SYNC_EVENT
0
用于打开或者关闭同步事件的监控。该监控项的生效必须是在ENABLE_MONITOR打开的情况下。1:打开;0关闭。该参数涉及到的动态视图为V S Y S T E M E V E N T 、 V SYSTEM_EVENT、V SYSTEMEVENTVSESSION_EVENT

MONITOR_SQL_EXEC
0
用于打开或者关闭操作符、困拟机栈帧、执行计划节点的监控。该监控项的生效必须是在ENABLE MONTOR.数涉及的动态规图为VSTKRM、VSSQL LANNODE、VSQL NODE HISTOR"

ENABLE_FREQROOTS
O
是否启用常用B树地址收集。0:不启用;1:启用

MAX_FREQ_ROOTS
200000
收集常用B树地址的最大数量;仅在ENABLE_FREQROOTS为1时有效,有效值范围(O~10000000)

MIN_FREQ_CNT
100000
最常使用B树地址的阈值,仅在ENABLE_FREQROOTS为1时有效,有效值范围(O~10000000)

LARGE_MEM_THRESHOLD
1000
大内存监控阈值。单位K,取值范围(O10000000)。其中O100关闭统计,100以上才统计。

ENABLE_MONITOR_DMSQL
1
启用动态监控SQL执行时间功能标记,0:不启用;1:监控。只监控多条或复杂SQL

ENABLE_TIMER_TRIG_LOG
0
ENABLE_TIMER_TRIG_LOG

IO_TIMEOUT
60
IO_TIMEOUT

数据守护相关参数

DW_MAX_SVR_WAIT_TIME
0
服务器等待守护进程启动的最大时间(秒)。取值范围(0~65534)。当此参数的值不为O时,实例…动的时间超过设置的值,则实例自动退出
,当参数的值等于0时,代表服务器不检测守护进程有没有启动

DW_INACTIVE_INTERVAL
60
DW_INACTIVE_INTERVAL

DW_PORT
0
服务器们守护进程之间的TCP通信徜口服务影监听此姝口接收守护进程的TCP洼接清求。有效苑围(102…没有配置TCP通信方式的守护进程不允许和DW UDP PORTINST UDP PORT同时配置

ALTER_MODE_STATUS
1
是否允许手工修改服务器的模式和状态,1表示允许A表示不允许默认为1。数据守护环境下建议配置为0,实例处于主机或备机模式后不允许用户直接通过SQL语句修改服务器的模式和状态

ENABLE_OFFLINE_TS
1
是否允许OFFLINE表空间,1表示允许,0表示不允许,2表示禁止备库,其他放开

SESS_FREE_IN_SUSPEND
60
远程归档失败会导致系统挂起,为了防止应用的连接一直挂住不切换到新主库;设置该参数,表示归档失败挂起后隔一段时间自动断开所有
连接。默认值60S,取值范围0-1800S,0表示不断开

SUSPEND_WORKER_TIMEOUT
600
挂起线程超时时间

SUSPENDING_FORBIDDEN
0
SUSPENDING_FORBIDDEN

DW_CONSISTENCY_CHECK
0
DW_CONSISTENCY_CHECK

全文索引相关参数

CTI_HASH_SIZE
100000
使用全文索引查询时,用于设置关键字匹配的哈希表大小。有效值范围(1000~10000000)

CTI_HASH_BUF_SIZE
50
使用全文索引查询时,用于设置哈希缓存内存大小,以M为单位。有效值范围(O~4000)

USE_FORALL_ATTR
0
是否使用FORALL语句的游标属性,0:不使用;1:使用

其他

ALTER_TABLE_OPT
0
是否对加列、修改列、删除列操作进行优化,0:全部不优化;1:全部优化;2:打开快速加列,对于删除列和修改列与1等效

IDLE_MEM_THRESHOLD
50
可用物理内存的报瞽阀值,单位是兆(10~6000)

IDLE_DISK_THRESHOLD
1000
磁盘可用空间的报警阀值,单位是兆(50~50000)

IDLE_SESS_THRESHOLD
5
有限的会话数阀值(1~10)

ENABLE_PRISVC
0
是否启用服务优先级的功能,0:不启用;1:启用

DSC_N_CTLS
10000
DSC_N_CTLS

DSC_N_POOLS
2
DSC_N_POOLS

DSC_USE_SBT
1
DSC_USE_SBT

DSC_TRX_CMT_LSN_SYNC
3
DSC_TRX_CMT_LSN_SYNC

DSC_ENABLE_MONITOR
1
DSC_ENABLE_MONITOR

DSC_GBS_REVOKE_OPT
1
DSC_GBS_REVOKE_OPT

DSC_HALT_SYNC
0
DSC_HALT_SYNC

DSC_TRX_VIEW_SYNC
0
DSC_TRX_VIEW_SYNC

DSC_REMOTE_READ_MODE
1
DSC_REMOTE_READ_MODE

DSC_FILE_INIT_ASYNC
1
DSC_FILE_INIT_ASYNC

DSC_TRX_CMT_OPT
0
DSC_TRX_CMT_OPT

DSC_RESERVE_PERCENT
0.080
DSC_RESERVE_PERCENT

DCP相关参数

ENABLE_DCP_MODE
0
服务器是否作为DCP集群代理运行,0:否;1:是

DCP_PORT_NUM
5237
DCP管理端口号

DCP_CONN_POOL_SIZE
1000
DCP连接池大小,有效值范围(1~100000)

USE_RDMA
0
USE_RDMA

配置文件相关参数

MAL_INI
0
是否启用MAIL系统,0:不启用;1:启用

ARCH_INI
1
是否启用归档,0:不启用;1:启用

REP_INI
0
是否启用复制,0:不启用;1:启用

LLOG_INI
0
是否启用逻辑日志,0:不启用;1:数据复制使用

TIMER_INI
0
是否启用定时器,0:不启用;1:启用

MPP_INI
0
是否启用MPP系统,0:不启用;1:启用

DFS_INI
0
DFS_INI

DSC相关参数

HA_INST_CHECK_IP

HA实例启动检测IP,是指HA系统中另外一台机器的IP地址,考虑到多网卡情况,最多允许配置5个IP

HA_INST_CHECK_PORT
65534
HA实例监听端口,数据库实例启动后;监听此端口的连接请求,并使用此端口号连接HAINST_CHECK_IP判断另外一个HA实例是否已经启动

REDO_UNTIL_LSN

系统故障重启时,重做REDO日志的最大LSN值

英文版

#DaMeng Database Server Configuration file
#this is comments

#file location of dm.ctl
CTL_PATH = /mnt/data/DAMENG/dm.ctl #ctl file path
CTL_BAK_PATH = /mnt/data/DAMENG/ctl_bak #dm.ctl backup path
CTL_BAK_NUM = 10 #backup number of dm.ctl, allowed to keep one more backup file besides specified number.
SYSTEM_PATH = /mnt/data/DAMENG #system path
CONFIG_PATH = /mnt/data/DAMENG #config path
TEMP_PATH = /mnt/data/DAMENG #temporary file path
BAK_PATH = /mnt/data/DAMENG/bak #backup file path
DFS_PATH = $/DAMENG #path of db_file in dfs

#instance name
INSTANCE_NAME = DMSERVER #Instance name

#memory pool and buffer
MAX_OS_MEMORY = 100 #Maximum Percent Of OS Memory
MEMORY_POOL = 188 #Memory Pool Size In Megabyte
MEMORY_N_POOLS = 0 #Number of Memory Pool
MEMORY_TARGET = 0 #Memory Share Pool Target Size In Megabyte
MEMORY_EXTENT_SIZE = 1 #Memory Extent Size In Megabyte
MEMORY_LEAK_CHECK = 0 #Memory Pool Leak Checking Flag
MEMORY_MAGIC_CHECK = 1 #Memory Pool Magic Checking Flag
MEMORY_BAK_POOL = 4 #Memory Backup Pool Size In Megabyte
HUGE_MEMORY_PERCENTAGE = 50 #Maximum percent of HUGE buffer that can be allocated to work as common memory pool
HUGE_BUFFER = 80 #Initial Huge Buffer Size In Megabytes
HUGE_BUFFER_POOLS = 4 #number of Huge buffer pools
BUFFER = 1000 #Initial System Buffer Size In Megabytes
BUFFER_POOLS = 5 #number of buffer pools
FAST_POOL_PAGES = 10000 #number of pages for fast pool
FAST_ROLL_PAGES = 8000 #number of pages for fast roll pages
KEEP = 8 #system KEEP buffer size in Megabytes
RECYCLE = 120 #system RECYCLE buffer size in Megabytes
RECYCLE_POOLS = 2 #Number of recycle buffer pools
ROLLSEG = 1 #system ROLLSEG buffer size in Megabytes
ROLLSEG_POOLS = 19 #Number of rollseg buffer pools
MULTI_PAGE_GET_NUM = 1 #Maximum number of pages for each read of buffer
PRELOAD_SCAN_NUM = 0 #The number of pages scanned continuously to start preload task
PRELOAD_EXTENT_NUM = 0 #The number of clusters preloaded for the first time
SORT_BUF_SIZE = 10 #maximum sort buffer size in Megabytes
SORT_BLK_SIZE = 1 #maximum sort blk size in Megabytes
SORT_BUF_GLOBAL_SIZE = 500 #maximum global sort buffer size in Megabytes
SORT_FLAG = 0 #choose method of sort
HAGR_HASH_SIZE = 100000 #hash table size for hagr
HJ_BUF_GLOBAL_SIZE = 500 #maximum hash buffer size for all hash join in Megabytes
HJ_BUF_SIZE = 50 #maximum hash buffer size for single hash join in Megabytes
HJ_BLK_SIZE = 1 #hash buffer size allocated each time for hash join in Megabytes
HAGR_BUF_GLOBAL_SIZE = 500 #maximum buffer size for all hagr in Megabytes
HAGR_BUF_SIZE = 50 #maximum buffer size for single hagr in Megabytes
HAGR_BLK_SIZE = 1 #buffer size allocated each time for hagr in Megabytes
MTAB_MEM_SIZE = 8 #memory table size in Kilobytes
FTAB_MEM_SIZE = 0 #file table package size in Kilobytes
MMT_GLOBAL_SIZE = 4000 #memory map table global size in megabytes
MMT_SIZE = 0 #memory map table size in megabytes
MMT_FLAG = 1 #ways of storing bdta data in memory map table
DICT_BUF_SIZE = 50 #dictionary buffer size in Megabytes
HFS_CACHE_SIZE = 160 #hfs cache size in Megabytes, used in huge horizon table for insert, update,delete
VM_STACK_SIZE = 256 #VM stack size in Kilobytes
VM_POOL_SIZE = 64 #VM pool size in Kilobytes
VM_POOL_TARGET = 16384 #VM pool target size in Kilobytes
SESS_POOL_SIZE = 16 #session pool size in Kilobytes
SESS_POOL_TARGET = 16384 #session pool target size in Kilobytes
RT_HEAP_TARGET = 8192 #runtime heap target size in Kilobytes
VM_MEM_HEAP = 0 #Whether to allocate memory to VM from HEAP
RFIL_RECV_BUF_SIZE = 16 #redo file recover buffer size in Megabytes
N_MEM_POOLS = 1 #number of memory pools
COLDATA_POOL_SIZE = 0 #coldata pool size for each worker group
HAGR_DISTINCT_HASH_TABLE_SIZE = 10000 #Size of hagr distinct hash table
CNNTB_HASH_TABLE_SIZE = 100 #Size of hash table in connect-by operation
GLOBAL_RTREE_BUF_SIZE = 100 #The total size of buffer for rtree
SINGLE_RTREE_BUF_SIZE = 10 #The size of buffer for single rtree
SORT_OPT_SIZE = 0 #once max memory size of radix sort assist count array
DFS_BUF_FLUSH_OPT = 0 #Whether to flush buffer page in opt mode for DFS storage

#thread
WORKER_THREADS = 1 #Number Of Worker Threads
TASK_THREADS = 4 #Number Of Task Threads
UTHR_FLAG = 0 #User Thread Flag
FAST_RW_LOCK = 1 #Fast Read Write Lock flag
SPIN_TIME = 4000 #Spin Time For Threads In Microseconds
WORK_THRD_STACK_SIZE = 8192 #Worker Thread Stack Size In Kilobytes
WORK_THRD_RESERVE_SIZE = 100 #Worker Thread Reserve Stack Size In Kilobytes
WORKER_CPU_PERCENT = 0 #Percent of CPU number special for worker thread
NESTED_C_STYLE_COMMENT = 0 #flag for C stype nested comment

#query
USE_PLN_POOL = 1 #Query Plan Reuse Mode, 0: Forbidden; 1:strictly reuse, 2:parsing reuse, 3:mixed parsing reuse
DYN_SQL_CAN_CACHE = 1 #Dynamic SQL cache mode. 0: Forbidden; 1: Allowed if the USE_PLN_POOL is non-zero;
VPD_CAN_CACHE = 0 #VPD SQL cache mode. 0: Forbidden; 1: Allowed if the USE_PLN_POOL is non-zero;
RS_CAN_CACHE = 0 #Resultset cache mode. 0: Forbidden; 1: Allowed only if the USE_PLN_POOL is non-zero;
RS_CACHE_TABLES = #Tables allowed to enable result set cache
RS_CACHE_MIN_TIME = 0 #Least time for resultset to be cached
RS_BDTA_FLAG = 0 #Resultset mode. 0: row; 2: bdta;
RS_BDTA_BUF_SIZE = 32 #Maximum size of message in Kilobytes for BDTA cursor, it’s valid only if RS_BDTA_FLAG is set to 2
RESULT_SET_LIMIT = 10000 #Maximum Number Of cached Resultsets
RESULT_SET_FOR_QUERY = 0 #Whether to generate result set for non-query statement
SESSION_RESULT_SET_LIMIT = 10000 #Maximum number of cached result sets for each session, 0 means unlimited
BUILD_FORWARD_RS = 0 #Whether to generate result set for forward only cursor
MAX_OPT_N_TABLES = 6 #Maximum Number Of Tables For Query Optimization
MAX_N_GRP_PUSH_DOWN = 5 #Maximum Number Of Rels For Group push down Optimization
CNNTB_MAX_LEVEL = 20000 #Maximum Level Of Hierarchical Query
CTE_MAXRECURSION = 100 #Maximum recursive Level Of Common Expression Table
CTE_OPT_FLAG = 1 #Optimize recursive with, 0: false, 1: convert refed subquery to invocation
BATCH_PARAM_OPT = 0 #optimize flag for DML with batch binded params
CLT_CONST_TO_PARAM = 0 #Whether to convert constant to parameter
LIKE_OPT_FLAG = 31 #the optimized flag of LIKE expression
FILTER_PUSH_DOWN = 0 #whether push down filter to base table
USE_MCLCT = 2 #mclct use flag for replace mgat
PHF_NTTS_OPT = 1 #phf ntts opt flag
MPP_MOTION_SYNC = 200 #mpp motion sync check number
UPD_DEL_OPT = 2 #update&delete opt flag, 0: false, 1: opt, 2: opt & ntts opt
ENABLE_INJECT_HINT = 0 #enable inject hint
FETCH_PACKAGE_SIZE = 512 #command fetch package size
UPD_QRY_LOCK_MODE = 0 #lock mode of FOR UPDATE query
ENABLE_DIST_IN_SUBQUERY_OPT = 0 #Whether to enable in-subquery optimization
MAX_OPT_N_OR_BEXPS = 7 #maximum number of OR bool expressions for query optimization
USE_HAGR_FLAG = 0 #Whether to use HAGR operator when can’t use SAGR operator
DTABLE_PULLUP_FLAG = 1 #the flag of pulling up derived table
VIEW_PULLUP_FLAG = 1 #the flag of pulling up view
GROUP_OPT_FLAG = 4 #the flag of opt group
FROM_OPT_FLAG = 0 #the flag of opt from
HAGR_PARALLEL_OPT_FLAG = 0 #the flag of opt hagr in mpp or parallel
HAGR_DISTINCT_OPT_FLAG = 2 #the flag of opt hagr distinct in mpp
REFED_EXISTS_OPT_FLAG = 1 #Whether to optimize correlated exists-subquery into non-correlated in-subquery
REFED_OPS_SUBQUERY_OPT_FLAG = 0 #Whether to optimize correlated op all/some/all-subquery into exists-subquery
HASH_PLL_OPT_FLAG = 0 #the flag of cutting partitioned table when used hash join
PARTIAL_JOIN_EVALUATION_FLAG = 1 #Whether to convert join type when upper operator is DISTINCT
USE_FK_REMOVE_TABLES_FLAG = 1 #Whether to remove redundant join by taking advantage of foreign key constraint
USE_FJ_REMOVE_TABLE_FLAG = 0 #Whether to remove redundant join by taking advantage of filter joining
SLCT_ERR_PROCESS_FLAG = 0 #How to handle error when processing single row
MPP_HASH_LR_RATE = 10 #The ratio of left child’s cost to right child’s cost of hash join in MPP environment that can influence the execution plan
LPQ_HASH_LR_RATE = 30 #The ratio of left child’s cost to right child’s cost of hash join in LPQ environment that can influence the execution plan
USE_HTAB = 1 #Whether to use HTAB operator for the whole plan
SEL_ITEM_HTAB_FLAG = 0 #Whether to use HTAB operator for correlated subquery in select items
OR_CVT_HTAB_FLAG = 1 #Whether to use HTAB operator to optimizer or-expression
ENHANCED_SUBQ_MERGING = 0 #Whether to use merging subquery opt
CASE_WHEN_CVT_IFUN = 9 #Flag of converting subquery in case-when expression to IF operator
OR_NBEXP_CVT_CASE_WHEN_FLAG = 0 #Whether to convert or-expression to case-when expression
NONCONST_OR_CVT_IN_LST_FLAG = 0 #Whether to convert nonconst or-expression to in lst expression
OUTER_CVT_INNER_PULL_UP_COND_FLAG = 1 #Whether to pull up join condition when outer join converts to inner join
OPT_OR_FOR_HUGE_TABLE_FLAG = 0 #Whether to use HFSEK to optimize or-expression for HUGE table
ORDER_BY_NULLS_FLAG = 0 #Whether to place NULL values to the end of the result set when in ascending order
SUBQ_CVT_SPL_FLAG = 1 #Flag of indicating how to convert correlated subquery
ENABLE_RQ_TO_SPL = 1 #Whether to convert correlated subquery to SPOOL
MULTI_IN_CVT_EXISTS = 0 #Whether to convert multi-column-in subquery to exists subquery
PRJT_REPLACE_NPAR = 1 #Whether to replace NPAR tree in NSEL after projection
ENABLE_RQ_TO_INV = 0 #Whether to convert correlated subquery to temporary function
SUBQ_EXP_CVT_FLAG = 0 #whether convert refered subquery exp to non-refered subquery exp
USE_REFER_TAB_ONLY = 0 #Whether to pull down correlated table only when dealing with correlated subquery
REFED_SUBQ_CROSS_FLAG = 1 #Whether to replace hash join with cross join for correlated subquery
IN_LIST_AS_JOIN_KEY = 0 #Whether to use in-list expression as join key
OUTER_JOIN_FLATING_FLAG = 0 #Flag of indicating whether outer join will be flattened
TOP_ORDER_OPT_FLAG = 0 #The flag of optimizing the query with the top cluase and the order by clause
PLACE_GROUP_BY_FLAG = 0 #The flag of optimizing the query with group_by and sfun by clause
TOP_DIS_HASH_FLAG = 1 #Flag of disable hash join in TOP-N query
ENABLE_RQ_TO_NONREF_SPL = 0 #Whether to convert correlated query to non-correlated query
OPTIMIZER_MODE = 1 #Optimizer_mode
NEW_MOTION = 0 #New Motion
LDIS_NEW_FOLD_FUN = 0 #ldis use different fold fun with mdis
DYNAMIC_CALC_NODES = 0 #different nodes of npln use different nubmer of calc sizes/threads
OPTIMIZER_MAX_PERM = 7200 #Optimizer_max permutations
ENABLE_INDEX_FILTER = 0 #enable index filter
OPTIMIZER_DYNAMIC_SAMPLING = 0 #Dynamic sampling level
TABLE_STAT_FLAG = 0 #How to use stat of table
AUTO_STAT_OBJ = 0 #Flag of automatically collecting statistics and recording DML changing rows
NONREFED_SUBQUERY_AS_CONST = 0 #Whether to convert non-correlated subquery to const
HASH_CMP_OPT_FLAG = 0 #Flag of operators that enable optimization with static hash table
OUTER_OPT_NLO_FLAG = 0 #Flag of enable index join for those whose right child is not base table
DISTINCT_USE_INDEX_SKIP = 2 #Distinct whether to use index skip scan
USE_INDEX_SKIP_SCAN = 0 #Whether to use index skip scan
INDEX_SKIP_SCAN_RATE = 0.003 #Rate in index skip scan
SPEED_SEMI_JOIN_PLAN = 1 #Flag of speeding up the generating process of semi join plan
COMPLEX_VIEW_MERGING = 0 #Flag of merging complex view into query without complex view
HLSM_FLAG = 1 #Choose one method to realize hlsm operator
DEL_HP_OPT_FLAG = 0 #Optimize delete for horization partition table
OPTIMIZER_OR_NBEXP = 0 #Flag of or-expression optimization method
CNNTB_OPT_FLAG = 0 #Optimize hierarchical query
ADAPTIVE_NPLN_FLAG = 0 #Adaptive npln
MULTI_UPD_OPT_FLAG = 0 #Optimize multi column update
MULTI_UPD_MAX_COL_NUM = 128 #Max value of column counts when optimize multi column update
ENHANCE_BIND_PEEKING = 0 #Enhanced bind peeking
NBEXP_OPT_FLAG = 3 #Whether to enable optimization for bool expressions
HAGR_HASH_ALGORITHM_FLAG = 0 #HAGR hash algorithm choice
DIST_HASH_ALGORITHM_FLAG = 0 #Distinct hash algorithm choice
UNPIVOT_ORDER_FLAG = 0 #Whether to sort the result of UNPIVOT operator by columns that don’t appear in the UNPIVOT sub-sentence
VIEW_FILTER_MERGING = 10 #Flag of merging view filter
ENABLE_PARTITION_WISE_OPT = 0 #whether enable partition-wise optimization
OPT_MEM_CHECK = 0 #reduce search space when out of memory
ENABLE_JOIN_FACTORIZATION = 0 #Whether to enable join factorization
EXPLAIN_SHOW_FACTOR = 1 #factor of explain
ERROR_COMPATIBLE_FLAG = 0 #enable/disable specified errors to be compatible with previous version
ENABLE_NEST_LOOP_JOIN_CACHE = 0 #whether enable cache temporary result of nest loop join child

#checkpoint
CKPT_RLOG_SIZE = 0 #Checkpoint Rlog Size, 0: Ingore; else: Generate With Redo Log Size
CKPT_DIRTY_PAGES = 0 #Checkpoint Dirty Pages, 0: Ingore; else: Generate With Dirty Pages
CKPT_INTERVAL = 300 #Checkpoint Interval In Seconds
CKPT_FLUSH_RATE = 5.00 #Checkpoint Flush Rate(0.0-100.0)
CKPT_FLUSH_PAGES = 1000 #Minimum number of flushed pages for checkpoints
CKPT_WAIT_PAGES = 128 #Maximum number of pages flushed for checkpoints
FORCE_FLUSH_PAGES = 0 #number of periodic flushed pages

#IO
DIRECT_IO = 0 #Flag For Io Mode(Non-Windows Only), 0: Using File System Cache; 1: Without Using File System Cache
IO_THR_GROUPS = 2 #The Number Of Io Thread Groups(Non-Windows Only)
HIO_THR_GROUPS = 2 #The Number Of Huge Io Thread Groups(Non-Windows Only)
FAST_EXTEND_WITH_DS = 1 #How To Extend File’s Size (Non-Windows Only), 0: Extend File With Hole; 1: Extend File With Disk Space

#database
MAX_SESSIONS = 1500 #Maximum number of concurrent sessions
MAX_CONCURRENT_TRX = 0 #Maximum number of concurrent transactions
CONCURRENT_TRX_MODE = 0 #Concurrent transactions mode
TRX_VIEW_SIZE = 512 #The buffer size of local transaction ids in TRX_VIEW
MAX_SESSION_STATEMENT = 20000 #Maximum number of statement handles for each session
MAX_SESSION_MEMORY = 0 #Maximum memory(In Megabytes) a single session can use
MAX_CONCURRENT_OLAP_QUERY = 0 #Maximum number of concurrent OLAP queries
BIG_TABLE_THRESHHOLD = 1000 #Threshhold value of a big table in 10k
MAX_EP_SITES = 64 #Maximum number of EP sites for MPP
PORT_NUM = 5236 #Port number on which the database server will listen
LISTEN_IP = #IP address from which the database server can accept
FAST_LOGIN = 0 #Whether to log information without login
DDL_AUTO_COMMIT = 1 #ddl auto commit mode, 0: not auto commit; 1: auto commit
COMPRESS_MODE = 0 #Default Compress Mode For Tables That Users Created, 0: Not Compress; 1: Compress
PK_WITH_CLUSTER = 0 #Default Flag For Primary Key With Cluster, 0: Non-Cluster; 1: Cluster
EXPR_N_LEVEL = 200 #Maximum nesting levels for expression
N_PARSE_LEVEL = 100 #Maximum nesting levels for parsing object
MAX_SQL_LEVEL = 500 #Maximum nesting levels of VM stack frame for sql
BDTA_SIZE = 16 #batch data processing size.SIZE OF BDTA(1-10000)
OLAP_FLAG = 2 #OLAP FLAG, 1 means enable olap
JOIN_HASH_SIZE = 500000 #the hash table size for hash join
HFILES_OPENED = 256 #maximum number of files can be opened at the same time for huge table
ISO_IGNORE = 0 #ignore isolation level flag
TEMP_SIZE = 1024 #temporary file size in Megabytes
TEMP_SPACE_LIMIT = 102400 #temp space limit in megabytes
FILE_TRACE = 0 #Whether to log operations of database files
COMM_TRACE = 1 #Whether to log warning information of communication
ERROR_TRACE = 0 #Whether to log error information, 1: NPAR ERROR
CACHE_POOL_SIZE = 200 #SQL buffer size in megabytes
PLN_DICT_HASH_THRESHOLD = 20 #Threshold in megabytes for plan dictionary hash table creating
STAT_COLLECT_SIZE = 10000 #minimum collect size in rows for statistics
STAT_ALL = 0 #if collect all the sub-tables of a partition table
PHC_MODE_ENFORCE = 0 #join mode
ENABLE_HASH_JOIN = 1 #enable hash join
ENABLE_INDEX_JOIN = 1 #enable index join
ENABLE_MERGE_JOIN = 1 #enable merge join
MPP_INDEX_JOIN_OPT_FLAG = 0 #enhance index inner join in mpp
MPP_NLI_OPT_FLAG = 0 #enhance nest loop inner join in mpp
MAX_PARALLEL_DEGREE = 1 #Maximum degree of parallel query
PARALLEL_POLICY = 0 #Parallel policy
PARALLEL_THRD_NUM = 10 #Thread number for parallel task
PARALLEL_MODE_COMMON_DEGREE = 1 #the common degree of parallel query for parallel-mode
PUSH_SUBQ = 0 #Whether to push down semi join for correlated subquery
OPTIMIZER_AGGR_GROUPBY_ELIM = 1 #Whether to attemp to eliminate group-by aggregations
UPD_TAB_INFO = 0 #Whether to update table info when startup
ENABLE_IN_VALUE_LIST_OPT = 6 #Flag of optimization methods for in-list expression
ENHANCED_BEXP_TRANS_GEN = 3 #Whether to enable enhanced transitive closure of boolean expressions
ENABLE_DIST_VIEW_UPDATE = 0 #whether view with distinct can be updated
STAR_TRANSFORMATION_ENABLED = 0 #Whether to enable star transformation for star join queries
MONITOR_INDEX_FLAG = 0 #monitor index flag
AUTO_COMPILE_FLAG = 1 #Whether to compile the invalid objects when loading
RAISE_CASE_NOT_FOUND = 0 #Whether raise CASE_NOT_FOUND exception for no case item matched
FIRST_ROWS = 100 #maximum number of rows when first returned to clients
LIST_TABLE = 0 #Whether to convert tables to LIST tables when created
ENABLE_SPACELIMIT_CHECK = 1 #flag for the space limit check, 0: disable 1: enable
BUILD_VERTICAL_PK_BTREE = 0 #Whether to build physical B-tree for primary key on huge tables
BDTA_PACKAGE_COMPRESS = 0 #Whether to compress BDTA packages
HFINS_PARALLEL_FLAG = 0 #Flag of parallel policy for inserting on huge table
HFINS_MAX_THRD_NUM = 100 #Maximum number of parallel threads that responsible for inserting on huge table
LINK_CONN_KEEP_TIME = 15 #Max idle time in minute for DBLINK before being closed
DETERMIN_CACHE_SIZE = 5 #deterministic function results cache size(M)
NTYPE_MAX_OBJ_NUM = 1000000 #Maximum number of objects and strings in composite data type
CTAB_SEL_WITH_CONS = 0 #Whether to build constraints when creating table by query
HLDR_BUF_SIZE = 8 #HUGE table fast loader buffer size in Megabytes
HLDR_BUF_TOTAL_SIZE = 4294967294 #HUGE table fast loader buffer total size in Megabytes
HLDR_REPAIR_FLAG = 0 #Flag of repairing huge table after exception, 0: NO 1: YES
HLDR_FORCE_COLUMN_STORAGE = 1 #Whether force column storage for last section data, 0: NO 1: YES
HLDR_FORCE_COLUMN_STORAGE_PERCENT = 80 #Minumun percent of unfully section data for huge force column storage
HLDR_HOLD_RATE = 1.50 #THE minimum rate to hold hldr of column number(1-65535)
HLDR_MAX_RATE = 2 #THE minimum rate to create hldr of column number(2-65535)
HUGE_ACID = 0 #Flag of concurrent mechanism for HUGE tables
HUGE_STAT_MODE = 2 #Flag of default stat mode when create huge table, 0:NONE 1:NORMAL 2:ASYNCHRONOUS
HFS_CHECK_SUM = 1 #Whether to check sum val for hfs data
HBUF_DATA_MODE = 0 #Whether to uncompress and decrypt data before read into HUGE buffer
DBLINK_OPT_FLAG = 509 #optimize dblink query flag
ELOG_REPORT_LINK_SQL = 0 #Whether to write the SQLs that sent to remote database by DBLINKs into error log file
FILL_COL_DESC_FLAG = 0 #Whether to return columns description while database returns results
BTR_SPLIT_MODE = 0 #Split mode for BTREE leaf, 0: split half and half, 1: split at insert point
BLOB_OUTROW_REC_STOR = 5 #Flag for blob outrow record storage
TS_RESERVED_EXTENTS = 64 #Number of reserved extents for each tablespace when startup
TS_SAFE_FREE_EXTENTS = 512 #Number of free extents which considered as safe value for each tablespace
TS_MAX_ID = 8192 #Maximum ID value for tablespaces in database
TS_FIL_MAX_ID = 255 #Maximum ID value for files in tablespace
DECIMAL_FIX_STORAGE = 0 #Whether convert decimal data to fixed length storage
SQL_SAFE_UPDATE_ROWS = 0 #Maximum rows can be effected in an update&delete statement
ENABLE_HUGE_SECIND = 1 #Whether support huge second index, 0: disable, 1: enable
TRXID_UNCOVERED = 0 #Whether disable scanning 2nd index only when pseudo column trxid used, 0: disable, 1: enable
LOB_MAX_INROW_LEN = 900 #Max lob data inrow len
RS_PRE_FETCH = 0 #Whether enable result pre-fetch
DFS_PAGE_FLUSH_VALIDATE = 1 #Whether enable validate page when flush in DDFS
GEN_SQL_MEM_RECLAIM = 1 #Whether reclaim memory space after generating each SQL’s plan
TIMER_TRIG_CHECK_INTERVAL = 60 #Server check timer trigger interval
DFS_HUGE_SUPPORT = 1 #Whether support huge table operation in DDFS
INNER_INDEX_DDL_SHOW = 1 #Whether to show inner index ddl.

#pre-load
LOAD_TABLE = #need to pre-load table
LOAD_HTABLE = #need to pre-load htable

#client cache
CLT_CACHE_TABLES = #Tables that can be cached in client

#redo log
RLOG_BUF_SIZE = 1024 #The Number Of Log Pages In One Log Buffer
RLOG_POOL_SIZE = 256 #Redo Log Pool Size In Megabyte
RLOG_PARALLEL_ENABLE = 0 #Whether to enable database to write redo logs in parallel mode
RLOG_IGNORE_TABLE_SET = 1 #Whether ignore table set
RLOG_APPEND_LOGIC = 0 #Type of logic records in redo logs
RLOG_APPEND_SYSTAB_LOGIC = 0 #Whether to write logic records of system tables in redo logs when RLOG_APPEND_LOGIC is set as 1
RLOG_SAFE_SPACE = 128 #Free redo log size in megabytes that can be considered as a save value
RLOG_SEND_APPLY_MON = 64 #Monitor recently send or apply rlog_buf info
RLOG_COMPRESS_LEVEL = 0 #The redo compress level,value in [0,10],0:do not compress
RLOG_ENC_CMPR_THREAD = 4 #The redo log thread number of encrypt and compress task,value in [1,64],default 4
RLOG_PKG_SEND_ECPR_ONLY = 0 #Only send encrypted or compressed data to standby instance without original data
DFS_RLOG_SEND_POLICY = 1 #DFS rlog send policy, 0: sync; 1:async

#redo redos
REDO_PWR_OPT = 1 #Whether to enable PWR optimization when system restarted after failure
REDO_IGNORE_DB_VERSION = 0 #Whether to check database version while database is redoing logs
REDO_BUF_SIZE = 64 #The max buffer size of rlog redo In Megabyte
REDOS_BUF_SIZE = 1024 #The max buffer size of rlog redo for standby In Megabyte
REDOS_MAX_DELAY = 1800 #The permitted max delay for one rlog buf redo on standby In Second
REDOS_BUF_NUM = 4096 #The max apply rlog buffer num of standby
REDOS_PRE_LOAD = 32 #Number of pre-load apply rlog buffer for standby
REDOS_PARALLEL_NUM = 1 #The parallel redo thread num

#transaction
ISOLATION_LEVEL = 1 #Default Transaction Isolation Level, 1: Read Commited; 3: Serializable
DDL_WAIT_TIME = 10 #Maximum waiting time in seconds for DDLs
BLDR_WAIT_TIME = 10 #Maximum waiting time in seconds for BLDR
MPP_WAIT_TIME = 10 #Maximum waiting time in seconds for locks on MPP
FAST_RELEASE_SLOCK = 1 #Whether to release share lock as soon as possible
SESS_CHECK_INTERVAL = 3 #Interval time in seconds for checking status of sessions
LOCK_TID_MODE = 1 #Lock mode for select-for-update operation
LOCK_TID_UPGRADE = 0 #Upgrade tid lock to X mode, 0:no, 1:yes
NOWAIT_WHEN_UNIQUE_CONFLICT = 0 #Whether to return immediately when unique constraint violation conflict happens
UNDO_EXTENT_NUM = 4 #Number of initial undo extents for each worker thread
MAX_DE_TIMEOUT = 10 #Maximum external function wait time in Seconds
TRX_RLOG_WAIT_MODE = 0 #Trx rlog wait mode
TRANSACTIONS = 75 #Maximum number of concurrent transactions
MVCC_RETRY_TIMES = 5 #Maximum retry times while MVCC conflicts happen
MVCC_PAGE_OPT = 1 #Page visible optimize for MVCC
ENABLE_FLASHBACK = 0 #Whether to enable flashback function
UNDO_RETENTION = 4800.000 #Maximum retention time in seconds for undo pages since relative transaction is committed
PARALLEL_PURGE_FLAG = 0 #flag for parallel purge of undo logs
PURGE_WAIT_TIME = 500 #Maximum wait time in microseconds for purging undo pages
PSEG_RECV = 3 #Whether to rollback active transactions and purge committed transactions when system restarts after failure
ENABLE_IGNORE_PURGE_REC = 0 #Whether to ignore purged records when returning -7120
ENABLE_TMP_TAB_ROLLBACK = 1 #enable temp table rollback
ROLL_ON_ERR = 0 #Rollback mode on Error, 0: rollback current statement 1: rollback whole transaction
XA_TRX_IDLE_TIME = 60 #Xa transaction idle time
LOB_MVCC = 1 #Whether LOB access in MVCC mode
LOCK_DICT_OPT = 2 #lock dict optimize
TRX_DICT_LOCK_NUM = 64 #Maximum ignorable dict lock number
ENABLE_ENCRYPT = 0 #Encrypt Mode For Communication, 0: Without Encryption; 1: SSL Encryption; 2: SSL Authentication
CLIENT_UKEY = 0 #Client ukey, 0: all, active by Client; 1: Force client ukey Authentication
MIN_SSL_VERSION = 771 #SSL minimum version For Communication, For example, 0: all, 0x0301: TLSv1, 0x0302: TLSv1.1, 0x0303: TLSv1.2, 0x0304: TLSv1.3
ENABLE_UDP = 0 #Enable udp For Communication, 0: disable; 1: single; 2: multi
UDP_MAX_IDLE = 15 #Udp max waiting time in second
UDP_BTU_COUNT = 8 #Count of udp batch transfer units
ENABLE_IPC = 0 #Enable ipc for communication, 0: disable; 1: enable
AUDIT_FILE_FULL_MODE = 1 #operation mode when audit file is full,1: delete old file; 2: no longer to write audit records
AUDIT_MAX_FILE_SIZE = 100 #maximum audit file size in Megabytes
AUDIT_IP_STYLE = 0 #IP style in audit record, 0: IP, 1: IP(hostname), default 0
MSG_COMPRESS_TYPE = 2 #Flag of message compression mode
LDAP_HOST = #LDAP Server ip
COMM_ENCRYPT_NAME = #Communication encrypt name, if it is null then the communication is not encrypted
COMM_VALIDATE = 1 #Whether to validate message
MESSAGE_CHECK = 0 #Whether to check message body
EXTERNAL_JFUN_PORT = 6363 #DmAgent port for external java fun.
EXTERNAL_AP_PORT = 4236 #DmAp port for external fun.
ENABLE_PL_SYNONYM = 0 #Whether try to resolve PL object name by synonym.
FORCE_CERTIFICATE_ENCRYPTION = 0 #Whether to encrypt login user name and password use certificate
REGEXP_MATCH_PATTERN = 0 #Regular expression match pattern, 0: support non-greedy match; 1: only support greedy match

#compatibility
BACKSLASH_ESCAPE = 0 #Escape Mode For Backslash, 0: Not Escape; 1: Escape
STR_LIKE_IGNORE_MATCH_END_SPACE = 1 #Whether to ignore end space of strings in like clause
CLOB_LIKE_MAX_LEN = 10240 #Maximum length in kilobytes of CLOB data that can be filtered by like clause
EXCLUDE_DB_NAME = #THE db names which DM7 server can exclude
MS_PARSE_PERMIT = 0 #Whether to support SQLSERVER’s parse style
COMPATIBLE_MODE = 0 #Server compatible mode, 0:none, 1:SQL92, 2:Oracle, 3:MS SQL Server, 4:MySQL, 5:DM6, 6:Teradata
DATETIME_FMT_MODE = 0 #Datetime fmt compatible mode, 0:none, 1:Oracle
DOUBLE_MODE = 0 #Calc double fold mode, 0:8bytes, 1:6bytes
CASE_COMPATIBLE_MODE = 1 #Case compatible mode, 0:none, 1:Oracle, 2:Oracle(new rule)
EXCLUDE_RESERVED_WORDS = #Reserved words to be exclude
COUNT_64BIT = 1 #Whether to set data type for COUNT as BIGINT
CALC_AS_DECIMAL = 0 #Whether integer CALC as decimal, 0: no, 1:only DIV, 2:all only has charactor
CMP_AS_DECIMAL = 0 #Whether integer compare as decimal, 0: no, 1:part, 2:all
CAST_VARCHAR_MODE = 1 #Whether to convert illegal string to special pseudo value when converting string to integer
PL_SQLCODE_COMPATIBLE = 0 #Whether to set SQLCODE in PL/SQL compatible with ORACLE as far as possible
LEGACY_SEQUENCE = 0 #Whether sequence in legacy mode, 0: no, 1:yes
DM6_TODATE_FMT = 0 #To Date’ HH fmt hour system, 0: 12(defalut) , 1: 24(DM6)
PK_MAP_TO_DIS = 0 #Whether map pk cols into distributed cols automatically
PROXY_PROTOCOL_MODE = 0 #PROXY PROTOCOL mode, 0: close; 1: open
DROP_CASCADE_VIEW = 0 #Whether to drop cascade view while dropping table or view

#request trace
SVR_LOG_NAME = SLOG_ALL #Using which sql log sys in sqllog.ini
SVR_LOG = 0 #Whether the Sql Log sys Is open or close. 1:open, 0:close, 2:use switch and detail mode. 3:use not switch and simple mode.

#system trace
GDB_THREAD_INFO = 0 #Generate gdb thread info while dm_sys_halt. 1: yes; 0: no
TRACE_PATH = #System trace path name
SVR_OUTPUT = 0 #Whether print at background

#ecs
ENABLE_ECS = 0 #Whether to enable elastic calc system
ECS_HB_INTERVAL = 60 #HB interval for elastic calc system
EP_GROUP_NAME = DMGROUP #Server group name as worker
AP_GROUP_NAME = DMGROUP #Server group name as cooperative worker
DCRS_IP = LOCALHOST #IP on which the dcrs server will listen
DCRS_PORT_NUM = 6236 #Port number on which the database dcrs will listen
AP_IP = LOCALHOST #The ap server IP for dcrs
AP_PORT_NUM = 0 #Port number on which the database ap will listen
AP_PARALLEL_DEGREE = 10 #Maximum degree of parallel for ap
ENABLE_ECS_MSG_CHECK = 0 #Whether to enable elastic check msg accumulate

#monitor
ENABLE_MONITOR = 1 #Whether to enable monitor
MONITOR_TIME = 1 #Whether to enable monitor timing
MONITOR_SYNC_EVENT = 0 #Whether to enable monitor sync event
MONITOR_SQL_EXEC = 0 #Whether to enable monitor sql execute
ENABLE_FREQROOTS = 0 #Whether to collect pages that used frequently
MAX_FREQ_ROOTS = 200000 #Maximum number of frequently used pages that will be collected
MIN_FREQ_CNT = 100000 #Minimum access counts of page that will be collected as frequently used pages
LARGE_MEM_THRESHOLD = 1000 #Large mem used threshold by k
ENABLE_MONITOR_DMSQL = 1 #Flag of performance monitoring:sql or method exec time.0: NO. 1: YES.
ENABLE_TIMER_TRIG_LOG = 0 #Whether to enable timer trigger log
IO_TIMEOUT = 300 #Maximum time in seconds that read from disk or write to disk

#data watch
DW_MAX_SVR_WAIT_TIME = 0 #Maximum time in seconds that server will wait for DMWATCHER to startup
DW_INACTIVE_INTERVAL = 60 #Time in seconds that used to determine whether DMWATCHER exist
DW_PORT = 0 #Instance tcp port for watch2
ALTER_MODE_STATUS = 1 #Whether to permit database user to alter database mode and status by SQLs, 1: yes, 0: no
ENABLE_OFFLINE_TS = 1 #Whether tablespace can be offline
SESS_FREE_IN_SUSPEND = 60 #Time in seconds for releasing all sessions in suspend mode after archive failed
SUSPEND_WORKER_TIMEOUT = 600 #Suspend worker thread timeout in seconds
SUSPENDING_FORBIDDEN = 0 #Suspending thread is forbidden
DW_CONSISTENCY_CHECK = 0 #Whether to check consistency for standby database, 1: yes, 0: no

#for context index
CTI_HASH_SIZE = 100000 #the hash table size for context index query
CTI_HASH_BUF_SIZE = 50 #the hash table cache size in Megabytes for context index query
USE_FORALL_ATTR = 0 #Whether to use cursor attributes of FORALL statements
ALTER_TABLE_OPT = 0 #Whether to optimize ALTER TABLE operation(add, modify or drop column)

#dcp
ENABLE_DCP_MODE = 0 #Whether to work as a proxy for database cluster
DCP_PORT_NUM = 5237 #Port number on which DCP will listen
DCP_CONN_POOL_SIZE = 1000 #DCP connection pool size
USE_RDMA = 0 #Whether to use rdma

#configuration file
MAL_INI = 0 #dmmal.ini
ARCH_INI = 0 #dmarch.ini
REP_INI = 0 #dmrep.ini
LLOG_INI = 0 #dmllog.ini
TIMER_INI = 0 #dmtimer.ini
MPP_INI = 0 #dmmpp.ini
DFS_INI = 0 #dmdfs.ini

#other
IDLE_MEM_THRESHOLD = 50 #minimum free memory warning size in Megabytes
IDLE_DISK_THRESHOLD = 1000 #minimum free disk space warning size in Megabytes
IDLE_SESS_THRESHOLD = 5 #minimum available session threshold value
ENABLE_PRISVC = 0 #Whether to enable priority service
DSC_N_CTLS = 10000 #Number Of LBS/GBS ctls
DSC_N_POOLS = 2 #Number Of LBS/GBS pools
DSC_USE_SBT = 1 #Use size balanced tree
DSC_TRX_CMT_LSN_SYNC = 3 #Whether to adjust lsn when trx commit
DSC_ENABLE_MONITOR = 1 #Whether to monitor request time
DSC_GBS_REVOKE_OPT = 1 #Whether to optimize GBS revoke
DSC_HALT_SYNC = 0 #Whether to halt other ep when resid_slot_wait overtime
DSC_TRX_VIEW_SYNC = 0 #Whether to wait response after broadcast trx view to other ep
DSC_REMOTE_READ_MODE = 1 #PAGE remote read optimize mode
DSC_FILE_INIT_ASYNC = 1 #DSC file init async flag
DSC_TRX_CMT_OPT = 0 #Whether to wait rlog file flush when trx commit in DSC
DSC_RESERVE_PERCENT = 0.080 #Start ctl reserve percent
HA_INST_CHECK_IP = #HA instance check IP
HA_INST_CHECK_PORT = 65534 #HA instance check port
PWR_FLUSH_PAGES = 1000 #Make special PWR rrec when n pages flushed
REDO_UNTIL_LSN = #redo until lsn
IGNORE_FILE_SYS_CHECK = 1 #ignore file sys check while starup
FILE_SCAN_PERCENT = 100.00 #percent of data scanned when calculating file used space
STARTUP_CHECKPOINT = 0 #checkpoint immediately when startup after redo
CHECK_SVR_VERSION = 1 #Whether to check server version
ID_RECYCLE_FLAG = 0 #Enable ID recycle
BAK_USE_AP = 1 #backup use assistant plus-in, 1:use AP; 2:not use AP. default 1.
ENABLE_DBLINK_TO_INV = 0 #Whether to convert correlated subquery which has dblink to temporary function
ENABLE_BLOB_CMP_FLAG = 0 #Whether BLOB/TEXT types are allowed to be compared
ENABLE_CREATE_BM_INDEX_FLAG = 1 #Allow bitmap index to be created
CVIEW_STAR_WITH_PREFIX = 1 #Whether append prefix for star item when create view
ENABLE_ADJUST_NLI_COST = 1 #Whether adjust cost of nest loop inner join
ENABLE_SEQ_REUSE = 0 #Whether allow reuse sequence expressions

  • 1
    点赞
  • 0
    评论
  • 2
    收藏
  • 一键三连
    一键三连
  • 扫一扫,分享海报

©️2021 CSDN 皮肤主题: 1024 设计师:白松林 返回首页
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、C币套餐、付费专栏及课程。

余额充值