TimesTen 应用层数据库缓存学习：4. 仅仅读缓存
在运行本文样例前。首先先运行TimesTen 应用层数据库缓存学习：2. 环境准备中的操作。
Read-only Cache Group的概念
仅仅读缓存组例如以下图：
仅仅读缓存组(Read-Only Cache Group)的cache table不能直接改动。在Oracle数据库中的数据更新后自己主动同步到Timesten。
仅仅读缓存的原理是对Oracle中须要缓存的表建立触发器来捕捉其变化
For an autorefresh cache group, TimesTen creates a change log table and trigger in the Oracle database for each cache table in the cache group. The trigger is fired for each committed insert, update, or delete operation on the cached Oracle Database table. The trigger records the primary key of the updated rows in the change log table. The cache agent periodically scans the change log table for updated keys and then joins this table with the cached Oracle Database table to get a snapshot of the latest updates.
cache table虽不能直接改动，但更新语句能够通过passthrough模式下发到Oracle运行。
假设TimesTen暂不可用。在Oracle中的更改能够暂存，待恢复正常后再同步到TimesTen。
如上图。仅仅读缓存组的操作通过cache agent实现，cache agent通过ttCacheUidPwdSet设置的username和口令连接到Oracle运行操作。
仅仅读缓存组不须要启动replication agent，仅仅有AWT缓存组才须要。
Read-only Cache Group的属性
从SQL Developer中，我们来看Read-only Cache Group的定义界面
Global表示是否为cache grid，read-only缓存组不支持，因此无法勾选。
选中Dynamic。表示载入方式为Dynamic Load，否则为Explicitly Load，以下的样例中会说明差别。
Autorefresh的属性中。incremental表示依据Oracle中记录的cached table的变化日志来同步，Full表示每次都unload全部数据后在全然同步。
Incremental autorefresh mode incurs some overhead to refresh the cache group for each committed update on the cached Oracle Database tables. There is no overhead when using full autorefresh mode.
上面这句的意思是。incremental autorefresh会在Oracle中记录日志，有开销。而full模式不会。可是。因为full模式是全同步。在refresh时的开销会非常大。而incremental模式在refresh时开销小。
假设每天夜间才须要同步一次。full模式也是合适的选择。
refresh interval是同步的周期。太密影响性能，太疏则数据实时性差。详细取决于业务需求。
手工load/refresh时，要求refresh state为paused。手工load/refresh和autorefresh完毕后，会将state改为on。
refresh state设置为off或paused时。autorefresh不会运行。
准备工作
在建立cache group之前，首先配置cache grid。
这里概念easy混淆，因为我们说cache group和cache grid是不同的。而这里的操作又是必需的。我们能够理解这里的配置工作为建立一个cache grid的框架，没有不论什么TimesTen节点attach到它上面。仅仅是便于以后扩展。
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=cacheadm;pwd=timesten;oraclepwd=oracle"
Command> call ttGridCreate ('samplegrid');
Command> call ttGridInfo;
< SAMPLEGRID, CACHEADM, Linux Intel x86, 32-bit, 11, 2, 2 >
Command> call ttGridNameSet ('samplegrid');
Command> call ttGridNameGet;
< SAMPLEGRID >然后启动cache agent。
$ ttisql -v1 cachedb1_1122
Command> call ttCacheStart;
或者
$ ttAdmin -cacheStart cachedb1_1122
然后能够通过ttStatus查看cache agent的状态：
$ ttstatus cachedb1_1122
TimesTen status report as of Sun Apr 10 06:24:24 2016
Daemon pid 2046 port 53392 instance tt1122
No TimesTen server running
------------------------------------------------------------------------
Data store /home/oracle/app/oracle/product/TimesTen/tt1122/info/DemoDataStore/cachedb1_1122
There are 14 connections to the data store
Shared Memory KEY 0x3a41077e ID 8650765
PL/SQL Memory KEY 0x3b41077e ID 8683534 Address 0x10000000
Type PID Context Connection Name ConnID
Cache Agent 17221 0x09a41200 Handler 2
Cache Agent 17221 0x09b70538 Timer 3
Cache Agent 17221 0xa7a0c560 Aging 4
Load和Refresh的差别
这时非常重要的概念，详述请參见Cache User Guide, Chapter 5, 5Cache Group Operations,Loading and refreshing a cache group
LOAD CACHE GROUP only loads committed inserts on the cached Oracle Database tables into the TimesTen cache tables. New cache instances are loaded into the cache tables, but cache instances that already exist in the cache tables are not updated or deleted even if the corresponding rows in the cached Oracle Database tables have been updated or deleted. A load operation is primarily used to initially populate a cache group.
REFRESH CACHE GROUP replaces cache instances in the TimesTen cache tables with the most current data from the cached Oracle Database tables including cache instances that are already exist in the cache tables. A refresh operation is primarily used to update the contents of a cache group with committed updates on the cached Oracle Database tables after the cache group has been initially populated.
If the cache table definitions use a WHERE clause, only rows that satisfy the WHERE clause are loaded or refreshed into the cache tables even if the LOAD CACHE GROUP or REFRESH CACHE GROUP statement does not use a WHERE clause.
To prevent a load or refresh operation from processing a large number of cache instances within a single transaction, which can greatly reduce concurrency and throughput, use the COMMIT EVERY n ROWS clause to specify a commit frequency unless you are using the WITH ID clause. If you specify COMMIT EVERY 0 ROWS, the load or refresh operation is processed in a single transaction.
Dynamic Read-Only Cache Group
$ cd $TT_HOME/quickstart/sample_scripts/cachegrid/
$ ls
create_global_awt.sql create_local_awt.sql create_local_dyn_ro.sql create_local_ro.sql
$ cat create_local_dyn_ro_bak.sql
create dynamic readonly cache group d_ro
autorefresh interval 1 second
from
tthr.jobs ( job_id varchar2(10) not null primary key,
job_title varchar2(35) not null,
min_salary number(6),
max_salary number(6));
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=cacheadm;pwd=timesten;oraclepwd=oracle" -f create_local_dyn_ro_bak.sql
5140: Could not find TTHR.JOBS in Oracle. May not have privileges.
$ sqlplus tthr/oracle@ttorcl
SQL*Plus: Release 11.2.0.2.0 Production on Sun Apr 10 06:45:32 2016
Copyright (c) 1982, 2010, Oracle. All rights reserved.
Connected to:
Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
SQL> select * from jobs;
JOB_ID JOB_TITLE MIN_SALARY MAX_SALARY
---------- ----------------------------------- ---------- ----------
AD_PRES President 20000 40000
AD_VP Administration Vice President 15000 30000
AD_ASST Administration Assistant 3000 6000
FI_MGR Finance Manager 8200 16000
FI_ACCOUNT Accountant 4200 9000
AC_MGR Accounting Manager 8200 16000
AC_ACCOUNT Public Accountant 4200 9000
SA_MAN Sales Manager 10000 20000
SA_REP Sales Representative 6000 12000
PU_MAN Purchasing Manager 8000 15000
PU_CLERK Purchasing Clerk 2500 5500
JOB_ID JOB_TITLE MIN_SALARY MAX_SALARY
---------- ----------------------------------- ---------- ----------
ST_MAN Stock Manager 5500 8500
ST_CLERK Stock Clerk 2000 5000
SH_CLERK Shipping Clerk 2500 5500
IT_PROG Programmer 4000 10000
MK_MAN Marketing Manager 9000 15000
MK_REP Marketing Representative 4000 9000
HR_REP Human Resources Representative 4000 9000
PR_REP Public Relations Representative 4500 10500
19 rows selected.
SQL> grant select on jobs to cacheadm;
Grant succeeded.
...
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=cacheadm;pwd=timesten;oraclepwd=oracle" -f create_local_dyn_ro_bak.sql
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=cacheadm;pwd=timesten"
Command> cachegroups;
Cache Group CACHEADM.D_RO:
Cache Group Type: Read Only (Dynamic)
Autorefresh: Yes
Autorefresh Mode: Incremental
Autorefresh State: Paused
Autorefresh Interval: 1 Second
Autorefresh Status: ok
Aging: LRU on
Root Table: TTHR.JOBS
Table Type: Read Only
# 在Oracle中，也能够看到缓存组的信息
$ cd $TT_HOME/oraclescripts
$ sqlplus cacheadm/oracle@ttorcl
SQL*Plus: Release 11.2.0.2.0 Production on Sun Apr 10 20:40:42 2016
Copyright (c) 1982, 2010, Oracle. All rights reserved.
Connected to:
Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
SQL> @cacheInfo.sql
*************Autorefresh Objects Information ***************
Host name: localhost.localdomain
Timesten datastore name:
/home/oracle/app/oracle/product/TimesTen/tt1122/info/DemoDataStore/cachedb1_1122
Cache table name: TTHR.JOBS
Change log table name: tt_06_114329_L
Number of rows in change log table: 1
Maximum logseq on the change log table: 1
Timesten has autorefreshed updates upto logseq: 1
Number of updates waiting to be autorefreshed: 0
Number of updates that has not been marked with a valid logseq: 0
****************************
*************No DDL Tracking objects are found*************
PL/SQL procedure successfully completed.
基本行为
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=tthr;pwd=timesten"
Command> desc jobs;
Table TTHR.JOBS:
Columns:
*JOB_ID VARCHAR2 (10) INLINE NOT NULL
JOB_TITLE VARCHAR2 (35) INLINE NOT NULL
MIN_SALARY NUMBER (6)
MAX_SALARY NUMBER (6)
Aging: LRU on
Command> select * from jobs;
Command> select * from jobs where job_id = 'FI_MGR';
5213: Bad Oracle login error in OCISessionBegin(): ORA-01017: invalid username/password; logon denied rc = -1
5131: Cannot connect to backend database: OracleNetServiceName = "TTORCL", uid = "TTHR", pwd is hidden, TNS_ADMIN = "/home/oracle/app/oracle/product/11.2.0/dbhome_2/network/admin", ORACLE_HOME= "/home/oracle/app/oracle/product/11.2.0/dbhome_2"
5109: Cache Connect general error: BDB connection not open.
Command> exit
# 这点特别不能理解，明明之前用ttCacheUidPwdSet设置了username和口令，但在进行dynamic load时，仍然须要指定oraclepwd才干成功。
# 看一下我的KM
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=tthr;pwd=timesten;oraclepwd=oracle"
# 以下的语句满足dynamic load要求，Dynamic Load每次仅仅载入一行数据，并且仅当数据和主键/唯一索引全然匹配时
Command> select * from jobs where job_id = 'FI_MGR';
< FI_MGR, Finance Manager, 8200, 16000 >
# 以下的语句不会引起dynamic load
Command> select * from jobs where MIN_SALARY = 3000;
Command> delete from jobs;
8225: Table JOBS is read only
#仅仅读缓存组不同意写通过passthrough模式訪问Oracle数据库
因为仅仅读缓存组不同意改动。我们可通过passthrough模式对Oracle进行操作
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=tthr;pwd=timesten;oraclepwd=oracle"
Command> select * from jobs;
< FI_MGR, Finance Manager, 8200, 16000 >
Command> passthrough 3
Command> select * from jobs;
< AD_PRES, President, 20000, 40000 >
< AD_VP, Administration Vice President, 15000, 30000 >
< AD_ASST, Administration Assistant, 3000, 6000 >
< FI_MGR, Finance Manager, 8200, 16000 >
< FI_ACCOUNT, Accountant, 4200, 9000 >
< AC_MGR, Accounting Manager, 8200, 16000 >
< AC_ACCOUNT, Public Accountant, 4200, 9000 >
< SA_MAN, Sales Manager, 10000, 20000 >
< SA_REP, Sales Representative, 6000, 12000 >
< PU_MAN, Purchasing Manager, 8000, 15000 >
< PU_CLERK, Purchasing Clerk, 2500, 5500 >
< ST_MAN, Stock Manager, 5500, 8500 >
< ST_CLERK, Stock Clerk, 2000, 5000 >
< SH_CLERK, Shipping Clerk, 2500, 5500 >
< IT_PROG, Programmer, 4000, 10000 >
< MK_MAN, Marketing Manager, 9000, 15000 >
< MK_REP, Marketing Representative, 4000, 9000 >
< HR_REP, Human Resources Representative, 4000, 9000 >
< PR_REP, Public Relations Representative, 4500, 10500 >
Command> delete from jobs where job_id = 'FI_MGR';
5224: Oracle foreign key violation error in OCIStmtExecute(): ORA-02292: integrity constraint (TTHR.EMP_JOB_FK) violated - child record found rc = -1
Command> update jobs set min_salary = 9000 where job_id = 'FI_MGR';
Command> commit;
Command> passthrough 0
Command> select * from jobs;
< FI_MGR, Finance Manager, 9000, 16000 >
在上例中，我们通过passthrough更改min_salary。然后更改同步到TimesTen
Dynamic Load 缓存组的Load和Refresh
以下的描写叙述仅针对Dynamic Load 缓存组
For a dynamic cache group, a refresh operation only refreshes committed updates and deletes on the cached Oracle Database tables into the cache tables because only existing cache instances in the cache tables are refreshed. New cache instances are not loaded into the cache tables so after the refresh operation completes, the cache tables contain either the same or fewer number of cache instances. To load new cache instances into the cache tables of a dynamic cache group, use a LOAD CACHE GROUP statement or perform a dynamic load operation. See “Dynamically loading a cache instance” for more information about a dynamic load operation.
Load和Refreh的最大差别在于, Load针对的是TimesTen缓存表中没有的记录。对于已有的记录不会影响。而Refresh仅仅更新TimesTen缓存表中已有的记录。
Load通经常使用于数据初始化载入，而Refresh用于更新缓存表中的数据，仅仅有当refresh interval较长时有意义。否则autorefresh会自己主动更新。
以下的操作中，为了方便，我们在timesten中将tthr.jobs的SELECT权限赋予了cacheadm，在生产环境中，请不要这样做。
$ ttisql -v1 cachedb1_1122
Command> grant select on tthr.jobs to cacheadm
先来看Load操作
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=cacheadm;pwd=timesten;oraclepwd=oracle"
Command> select * from tthr.jobs where job_id = 'FI_MGR';
< FI_MGR, Finance Manager, 9000, 16000 >
Command> load cache group CACHEADM.D_RO where min_salary > 10000 commit every 256 rows;
Command> select * from tthr.jobs;
< AD_PRES, President, 20000, 40000 >
< AD_VP, Administration Vice President, 15000, 30000 >
< FI_MGR, Finance Manager, 9000, 16000 >
# 这里尽管FI_MGR的最低工资并不是>10000。但仍然在结果集中，我们以下改动为20000
$ sqlplus tthr/oracle@ttorcl
SQL*Plus: Release 11.2.0.2.0 Production on Sun Apr 10 22:25:27 2016
Copyright (c) 1982, 2010, Oracle. All rights reserved.
Connected to:
Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
SQL> update jobs set min_salary = 20000 where job_id = 'FI_MGR';
1 row updated.
SQL> commit;
Commit complete.
# 回到ttisql会话
Command> load cache group CACHEADM.D_RO where min_salary > 10000 commit every 256 rows;
Command> select * from tthr.jobs;
< AD_PRES, President, 20000, 40000 >
< AD_VP, Administration Vice President, 15000, 30000 >
< FI_MGR, Finance Manager, 20000, 16000 >
# 奇怪，这回FI_MGR的min_salary更新了，但事实是，这不是因为load动作引起的。而是由autorefresh引起的
# 为了看的更清楚些。我们须要将refresh interval设置更长一些或者是停止autorefresh以避免autorefresh的干扰。我们先将其设置为1小时。
Command> unload cache group CACHEADM.D_RO;
Command> select count(*) from tthr.jobs;
< 0 >
Command> alter cache group CACHEADM.D_RO set autorefresh interval 3600 seconds;
Command> exit
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=tthr;pwd=timesten;oraclepwd=oracle"
Command> passthrough 3;
Command> update jobs set min_salary = 9000 where job_id = 'FI_MGR';
Command> commit;
Command> passthrough 0;
Command> select * from jobs;
Command> select * from jobs where job_id = 'FI_MGR';
< FI_MGR, Finance Manager, 9000, 16000 >
Command> exit
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=cacheadm;pwd=timesten;oraclepwd=oracle"
Command> load cache group CACHEADM.D_RO where min_salary > 10000 commit every 256 rows;
Command> select * from tthr.jobs;
< AD_PRES, President, 20000, 40000 >
< AD_VP, Administration Vice President, 15000, 30000 >
< FI_MGR, Finance Manager, 9000, 16000 >
# 在Oracle中将FI_MGR的min_salary改为20000
$ sqlplus tthr/oracle@ttorcl
SQL*Plus: Release 11.2.0.2.0 Production on Sun Apr 10 22:40:53 2016
Copyright (c) 1982, 2010, Oracle. All rights reserved.
Connected to:
Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
SQL> update jobs set min_salary = 20000 where job_id = 'FI_MGR';
1 row updated.
SQL> commit;
Commit complete.
SQL> exit
# 回到ttisql会话，再次load，因为refresh interval为1小时，此时还未发生，这时看到的结果没有autorefresh的影响。
# FI_MGR的min_salary仍不变。说明load仅仅载入满足条件，但不在cache table中的数据
Command> load cache group CACHEADM.D_RO where min_salary > 10000 commit every 256 rows;
Command> select * from tthr.jobs ;
< AD_PRES, President, 20000, 40000 >
< AD_VP, Administration Vice President, 15000, 30000 >
< FI_MGR, Finance Manager, 9000, 16000 >
Command> unload cache group CACHEADM.D_RO;再看Refresh操作，相同，我们须要避免autorefresh的干扰，这里我们将禁止掉autorefresh。
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=cacheadm;pwd=timesten;oraclepwd=oracle"
Command> unload cache group CACHEADM.D_RO;
Command> select * from tthr.jobs;
Command> select * from tthr.jobs where job_id = 'FI_MGR';
< FI_MGR, Finance Manager, 20000, 16000 >
# 禁止autorefresh
Command> alter cache group CACHEADM.D_RO set autorefresh state paused;
# 不同意带where条件
Command> refresh cache group CACHEADM.D_RO where min_salary > 10000 commit every 256 rows;
8304: Manual REFRESH operation with WHERE clause is not allowed on cache group CACHEADM.D_RO because it is specified as dynamic AUTOREFRESH
Command> refresh cache group CACHEADM.D_RO commit every 256 rows;
Command> select * from tthr.jobs;
< FI_MGR, Finance Manager, 20000, 16000 >
# 在Oracle中，将20000改为9000
$ sqlplus tthr/oracle@ttorcl
SQL*Plus: Release 11.2.0.2.0 Production on Sun Apr 10 23:04:40 2016
Copyright (c) 1982, 2010, Oracle. All rights reserved.
Connected to:
Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
SQL> update jobs set min_salary = 9000 where job_id = 'FI_MGR';
1 row updated.
SQL> commit;
Commit complete.
SQL> exit
Disconnected from Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
# 回到ttisql会话
Command> refresh cache group CACHEADM.D_RO commit every 256 rows;
8289: Manual LOAD/REFRESH of cache group requires AUTOREFRESH STATE to be PAUSED, but the STATE of CACHEADM.D_RO is ON. The LOAD is not allowed
# 前面refresh运行后。autorefresh会自己主动同意，因此我们仍需将其禁止
Command> alter cache group CACHEADM.D_RO set autorefresh state paused;
Command> refresh cache group CACHEADM.D_RO commit every 256 rows;
Command> select * from tthr.jobs;
< FI_MGR, Finance Manager, 9000, 16000 >**这里我们应该了解了refresh和load的差别，load相似于insert语句，仅仅载入timesten中没有的数据。而refresh相似于update。仅仅更新timesten中已有的数据。
autorefresh会触发refresh动作。**
Explicit Load Cache Group
$ cd $TT_HOME/quickstart/sample_scripts/cachegrid/
$ ls
create_global_awt.sql create_local_awt.sql create_local_dyn_ro.sql create_local_ro.sql
$ cat create_local_dyn_ro_bak.sql
$ cat create_local_ro.sql
create readonly cache group ro
autorefresh
interval 5 seconds
mode incremental
from
tthr.departments (
department_id number(4) not null primary key,
department_name varchar2(30) not null,
manager_id number(6),
location_id number(4))
where department_id < 100;
# 将hr.改动为tthr。加上where条件。
并另存为create_local_ro_bak.sql
$ sqlplus tthr/oracle@ttorcl
SQL*Plus: Release 11.2.0.2.0 Production on Sun Apr 10 23:42:55 2016
Copyright (c) 1982, 2010, Oracle. All rights reserved.
Connected to:
Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
SQL> select * from departments where department_id < 100;
DEPARTMENT_ID DEPARTMENT_NAME MANAGER_ID LOCATION_ID
------------- ------------------------------ ---------- -----------
10 Administration 200 1700
20 Marketing 201 1800
30 Purchasing 114 1700
40 Human Resources 203 2400
50 Shipping 121 1500
60 IT 103 1400
70 Public Relations 204 2700
80 Sales 145 2500
90 Executive 100 1700
9 rows selected.
SQL> grant select on departments to cacheadm;
Grant succeeded.
SQL> exit
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=cacheadm;pwd=timesten;oraclepwd=oracle" -f create_local_ro_bak.sql
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=cacheadm;pwd=timesten"
Command> cachegroups;
Cache Group CACHEADM.RO:
Cache Group Type: Read Only
Autorefresh: Yes
Autorefresh Mode: Incremental
Autorefresh State: Paused
Autorefresh Interval: 5 Seconds
Autorefresh Status: ok
Aging: No aging defined
Root Table: TTHR.DEPARTMENTS
Table Type: Read Only
# 在Oracle中，也能够看到缓存组的信息
$ cd $TT_HOME/oraclescripts
$ sqlplus cacheadm/oracle@ttorcl
SQL*Plus: Release 11.2.0.2.0 Production on Sun Apr 10 20:40:42 2016
Copyright (c) 1982, 2010, Oracle. All rights reserved.
Connected to:
Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
SQL> @cacheInfo.sql
*************Autorefresh Objects Information ***************
Host name: localhost.localdomain
Timesten datastore name:
/home/oracle/app/oracle/product/TimesTen/tt1122/info/DemoDataStore/cachedb1_1122
Cache table name: TTHR.DEPARTMENTS
Change log table name: tt_06_114326_L
Number of rows in change log table: 1
Maximum logseq on the change log table: 0
Timesten has autorefreshed updates upto logseq: -1
Number of updates waiting to be autorefreshed: 1
Number of updates that has not been marked with a valid logseq: 0
****************************
*************No DDL Tracking objects are found*************
PL/SQL procedure successfully completed.
基本行为
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=tthr;pwd=timesten"
Command> select * from departments;
Command> select * from departments where DEPARTMENT_ID = 10;
# 没有不论什么数据，须要手工load缓存组的Load和Refresh
以下的操作中，为了方便。我们在timesten中将tthr.departments的SELECT权限赋予了cacheadm。在生产环境中，请不要这样做。
$ ttisql -v1 cachedb1_1122
Command> grant select on tthr.departments to cacheadm
先来看Load
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=cacheadm;pwd=timesten;oraclepwd=oracle"
Command> load cache group cacheadm.ro commit every 256 rows;
Command> select count(*) from tthr.departments;
< 9 >
Command> unload cache group cacheadm.ro;
Command> select count(*) from tthr.departments;
< 0 >
Command> alter cache group CACHEADM.RO set autorefresh state paused;
# 对于非dynamic的仅仅读缓存组，load时不同意带where条件
Command> load cache group cacheadm.ro where department_id < 50 commit every 256 rows;
8287: Manual LOAD/RFRESH operations with WHERE or WITHID clauses are not allowed on cache group CACHEADM.RO because it is specified as AUTOREFRESH
Command> load cache group cacheadm.ro commit every 256 rows;
8289: Manual LOAD/REFRESH of cache group requires AUTOREFRESH STATE to be PAUSED, but the STATE of CACHEADM.RO is OFF. The LOAD is not allowed
Command> alter cache group CACHEADM.RO set autorefresh state paused;
Command> load cache group cacheadm.ro commit every 256 rows;
Command> select count(*) from tthr.departments;
< 9 >
# 在cache table非空时，不同意load
Command> load cache group cacheadm.ro commit every 256 rows;
8288: Manual LOAD of cache group CACHEADM.RO is not allowed because it is specified as AUTOREFRESH and it is not empty
再看Refresh
# 在Oracle中以tthr用户插入新记录
SQL> insert into departments values(25, 'SC', NULL, 1700);
SQL> commit;
# 在timesten中，立即能够看到这条记录
Command> select * from tthr.departments;
< 10, Administration, 200, 1700 >
< 20, Marketing, 201, 1800 >
**< 25, SC, <NULL>, 1700 >**
< 30, Purchasing, 114, 1700 >
< 40, Human Resources, 203, 2400 >
< 50, Shipping, 121, 1500 >
< 60, IT, 103, 1400 >
< 70, Public Relations, 204, 2700 >
< 80, Sales, 145, 2500 >
< 90, Executive, 100, 1700 >
# 我们没有手工运行不论什么Load和Refresh操作。而这条数据却同步到了TimesTen。这个结果是因为5秒一次的autorefresh导致的。
# 由此可见。Explicitly与Dynamic Load的缓存组，refresh的行为是不一样的，前者的refresh等于unload all + load all。而后者相当于update。
# 我们能够先禁止掉autorefresh后。然后看refresh的行为
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=cacheadm;pwd=timesten;oraclepwd=oracle"
Command> select * from tthr.departments;
< 10, Administration, 200, 1700 >
< 20, Marketing, 201, 1800 >
< 25, SC, <NULL>, 1700 >
< 30, Purchasing, 114, 1700 >
< 40, Human Resources, 203, 2400 >
< 50, Shipping, 121, 1500 >
< 60, IT, 103, 1400 >
< 70, Public Relations, 204, 2700 >
< 80, Sales, 145, 2500 >
< 90, Executive, 100, 1700 >
Command> alter cache group CACHEADM.RO set autorefresh state paused;
# 在Oracle中以tthr用户删除刚插入的记录
SQL> delete from departments where department_id = 25;
SQL> commit;
# 回到ttisql会话，因为autorefresh周期非常长，此时数据还在
Command> select * from tthr.departments where department_id = 25;
< 25, SC, <NULL>, 1700 >
# 手工refresh，这时数据同步了
Command> refresh cache group cacheadm.ro commit every 256 rows;
Command> select * from tthr.departments where department_id = 25;
For an explicitly loaded cache group, a refresh operation is equivalent to issuing an UNLOAD CACHE GROUP statement followed by a LOAD CACHE GROUP statement on the cache group. In effect, all committed inserts, updates and deletes on the cached Oracle Database tables are refreshed into the cache tables. New cache instances may be loaded into the cache tables. Cache instances that already exist in the cache tables are updated or deleted if the corresponding rows in the cached Oracle Database tables have been updated or deleted. See “Unloading a cache group” for more information about the UNLOAD CACHE GROUP statement.
能够看到，对于Explicitly load缓存组，数据初始化时的load操作是必须的。但也仅仅须要运行一次，手工refresh操作则相当于unload+load。开销非常大，而自己主动refresh操作会基于日志将Oracle中发生变化的记录同步到TimesTen。
大部分时候，refresh和load这些手工操作都是不必要的， TimesTen会自己主动的运行autorefresh，dynamic load等操作。
本文的样例仅仅为说明原理
仅仅读缓存组，缓存部分行和列
缓存部分列是在定义缓存组时实现的，比如：
CREATE READONLY CACHE GROUP "RO"
AUTOREFRESH MODE INCREMENTAL INTERVAL 5 MINUTES
STATE PAUSED
FROM
"TTHR"."DEPARTMENTS" (
"DEPARTMENT_ID" NUMBER(4) NOT NULL,
PRIMARY KEY("DEPARTMENT_ID")
)除了主键列必须缓存外，其他都能够不缓存
缓存部分行在定义缓存组时实现，比如
CREATE DYNAMIC READONLY CACHE GROUP "RO"
AUTOREFRESH MODE INCREMENTAL INTERVAL 5 MINUTES
STATE PAUSED
FROM
"TTHR"."DEPARTMENTS" (
"DEPARTMENT_ID" NUMBER(4) NOT NULL,
"DEPARTMENT_NAME" VARCHAR2(30 BYTE) NOT NULL,
"MANAGER_ID" NUMBER(6) ,
"LOCATION_ID" NUMBER(4) ,
PRIMARY KEY("DEPARTMENT_ID")
) AGING LRU OFF
WHERE (departments.department_id = 1)
Command> cachegroups ro
Cache Group CACHEADM.RO:
Cache Group Type: Read Only (Dynamic)
Autorefresh: Yes
Autorefresh Mode: Incremental
Autorefresh State: Paused
Autorefresh Interval: 5 Minutes
Autorefresh Status: ok
Aging: LRU off
Root Table: TTHR.DEPARTMENTS
Where Clause: (departments.department_id = 1)
Table Type: Read Only
root table和child table都能够加where条件：
CREATE READONLY CACHE GROUP "RO"
AUTOREFRESH MODE INCREMENTAL INTERVAL 5 MINUTES
STATE PAUSED
FROM
"TTHR"."REGIONS" (
"REGION_ID" NUMBER NOT NULL,
"REGION_NAME" VARCHAR2(25 BYTE),
PRIMARY KEY("REGION_ID")
)
WHERE (regions.region_id < 5),
"TTHR"."COUNTRIES" (
"COUNTRY_ID" CHAR(2 BYTE) NOT NULL,
"COUNTRY_NAME" VARCHAR2(40 BYTE),
"REGION_ID" NUMBER NOT NULL,
PRIMARY KEY("COUNTRY_ID"),
FOREIGN KEY("REGION_ID")
REFERENCES "TTHR"."REGIONS"("REGION_ID")
)
WHERE (tthr.countries.region_id < 5),
"TTHR"."LOCATIONS" (
"LOCATION_ID" NUMBER(4) NOT NULL,
"STREET_ADDRESS" VARCHAR2(40 BYTE),
"POSTAL_CODE" VARCHAR2(12 BYTE),
"CITY" VARCHAR2(30 BYTE) NOT NULL,
"STATE_PROVINCE" VARCHAR2(25 BYTE),
"COUNTRY_ID" CHAR(2 BYTE) NOT NULL,
PRIMARY KEY("LOCATION_ID"),
FOREIGN KEY("COUNTRY_ID")
REFERENCES "TTHR"."COUNTRIES"("COUNTRY_ID")
)
WHERE (locations.country_id > 0)缓存部分行是在Load时实现的，比如
load cache group "CACHEADM"."RO"
where department_id = 1
commit every 256 rows可是仅仅对dynamic read only 才有效
仅仅读缓存的限制
參考cache user guide, chapter 4.
The cache tables on TimesTen cannot be updated directly.
这不用解释了
Only the ON DELETE CASCADE and UNIQUE HASH ON cache table attributes can be used in the cache table definitions.A FLUSH CACHE GROUP statement cannot be issued on the cache group.
flush是从TimesTen到Oracle的手工动作，当然也不同意了
A TRUNCATE TABLE statement issued on a cached Oracle Database table is not automatically refreshed to the TimesTen cache table.
不会自己主动的更新。但能够通过refresh，或unload+load更新
A LOAD CACHE GROUP statement can only be issued on the cache group if the cache tables are empty, unless the cache group is dynamic.对于explicitly load缓存组，load时，cache table必须为空
The autorefresh state must be PAUSED before you can issue a LOAD CACHE GROUP statement on the cache group, unless the cache group is dynamic, in which case the autorefresh state must be PAUSED or ON.The LOAD CACHE GROUP statement cannot contain a WHERE clause, unless the cache group is dynamic, in which case the WHERE clause must be followed by a COMMIT EVERY n ROWS clause.
对于explicitly load缓存组，load时不能带where条件, 结合上一个条件。这时load仅仅能用于数据初始化The autorefresh state must be PAUSED before you can issue a REFRESH CACHE GROUP
statement on the cache group. The REFRESH CACHE GROUP statement cannot contain
a WHERE clauseAll tables and columns referenced in WHERE clauses when creating, loading or
unloading the cache group must be fully qualified.
清理cache group
$ ttisql -v1 -connstr "dsn=cachedb1_1122;uid=cacheadm;pwd=timesten;oraclepwd=oracle"
Command> call ttGridDetach;
3363: Node is not attached to a grid <-说明是cache group而非cache grid
Command> drop cache group CACHEADM.D_RO;
Command> drop cache group CACHEADM.RO;
Command> call ttcachestop;
Command> call ttGridDestroy('samplegrid');
Command> call ttGridNameGet;
< <NULL> >參考
- Cache User Guide, Chapter 2 - Getting Started，Performing operations on the read-only cache group
- Cache User Guide, Chapter 4 - Defining Cache Groups，Creating a cache group
- Cache User Guide, Chapter 5 - Cache Group Operations，Loading and refreshing a cache group
TimesTen 应用层数据库缓存学习：4. 仅仅读缓存的更多相关文章
- 使用Retrofit和Okhttp实现网络缓存。无网读缓存，有网根据过期时间重新请求 （转）
使用Retrofit和Okhttp实现网络缓存,更新于2016.02.02原文链接:http://www.jianshu.com/p/9c3b4ea108a7 本文使用 Retrofit2.0.0-b ...
- Java缓存学习之三：CDN缓存机制
CDN是什么? 关于CDN是什么,此前网友详细介绍过. CDN是Content Delivery Network的简称,即"内容分发网络"的意思.一般我们所说的CDN加速,一般是指 ...
- [原创]java WEB学习笔记93：Hibernate学习之路---Hibernate 缓存介绍，缓存级别，使用二级缓存的情况，二级缓存的架构集合缓存，二级缓存的并发策略，实现步骤，集合缓存，查询缓存，时间戳缓存
本博客的目的:①总结自己的学习过程,相当于学习笔记 ②将自己的经验分享给大家,相互学习,互相交流,不可商用 内容难免出现问题,欢迎指正,交流,探讨,可以留言,也可以通过以下方式联系. 本人互联网技术爱 ...
- Java缓存学习之二：浏览器缓存机制
浏览器端的九种缓存机制介绍 浏览器缓存是浏览器端保存数据用于快速读取或避免重复资源请求的优化机制,有效的缓存使用可以避免重复的网络请求和浏览器快速地读取本地数据,整体上加速网页展示给用户.浏览器端缓存 ...
- Java缓存学习之一：缓存
一.缓存 1.什么是缓存? 缓存是硬件,是CPU中的组件,CPU存取数据的速度非常的快,一秒钟能够存取.处理十亿条指令和数据(术语:CPU主频1G),而内存就慢很多,快的内存能够达到几十兆就不错了,可 ...
- RHCA442学习笔记-Unit11内存缓存
Unit 11 Memory Caches 内存缓存 学习目标: A. 使用内存来改善运行慢的子系统的服务时间. 11.1 Strategi ...
- MySql数据库3【优化3】缓存设置的优化
1.表缓存 相关参数: table_open_cache 指定表缓存的大小.每当MySQL访问一个表时,如果在表缓冲区中还有空间,该表就被打开并放入其中,这样可以更快地访问表内容.通过检查峰值时间的状 ...
- 闲话缓存：ZFS 读缓存深入研究-ARC（一）
在Solaris ZFS 中实现的ARC(Adjustable Replacement Cache)读缓存淘汰算法真是很有意义的一块软件代码.它是基于IBM的Megiddo和Modha提出的ARC(A ...
- 数据库MySQL学习笔记高级篇
数据库MySQL学习笔记高级篇 写在前面 学习链接:数据库 MySQL 视频教程全集 1. mysql的架构介绍 mysql简介 概述 高级Mysql 完整的mysql优化需要很深的功底,大公司甚至有 ...
随机推荐
- 【BZOJ 2744】 2744: [HEOI2012]朋友圈 (最大团，二分图匹配，构图)
2744: [HEOI2012]朋友圈 Description 在很久很久以前,曾经有两个国家和睦相处,无忧无虑的生活着.一年一度的评比大会开始了,作为和平的两国,一个朋友圈数量最多的永远都是最值得他 ...
- cogs 1075. [省常中2011S4] 最短路径问题
1075. [省常中2011S4] 最短路径问题 ★ 输入文件:short.in 输出文件:short.out 简单对比 时间限制:1 s 内存限制:128 MB [问题描述] 平面上 ...
- [BZOJ3997][TJOI2015]组合数学(Dilworth定理+DP)
题目名字是什么就不能往那方面想. 每个点拆成a[i][j]个,问题变为DAG最小路径覆盖,由Dilworth定理转成最长反链. 使用Dilworth定理的时候要注意那些点之间有边,这里任意一个点和其右 ...
- [拒绝毒瘤的小清新系列] give you a tree
题面在这里! (小声)这其实是我读错某题之后自己出的一道题...... 正解也很简单啦,直接扫描线+线段树水过(考虑一个合法的区间正好有 siz-1 对树上相邻的点对)23333 #include&l ...
- 【二分】Jessica's Reading Problem
[POJ3320]Jessica's Reading Problem Time Limit: 1000MS Memory Limit: 65536K Total Submissions: 1309 ...
- JS对象和数组
/* 数组和对象 [JavaScript 权威指南 第五版] */ /* 对象: 是一个无序属性集合, 每个属性都有自己的名字和值 */ /* 创建对象简单方法, 对象直接量 */ var obj = ...
- [转载]memcached 命令操作详解
转载:http://www.cnblogs.com/azheng007/p/3159345.html 一.存储命令 存储命令的格式: <command name> <key> ...
- Mock -- 数据模拟
作者:张云龙链接:https://www.zhihu.com/question/35436669/answer/62753889来源:知乎著作权归作者所有.商业转载请联系作者获得授权,非商业转载请注明 ...
- [Linux] Linux smaps接口文件结构
在Linux内核 2.6.16中引入了一个系统内存接口特性,这个接口位于/proc/$pid/目录下的smaps文件中 ,一看内容发现是进程内存映像信息,比同一目录下的maps文件更详细些. 400d ...
- Oracle之数据库安全
密码破解,大部分其实是通过枚举的方式,列出可能的密码,然后逐个尝试,直到找到真正的密码,有时也叫暴力破解.接下来,我们将举几个密码破解的例子. n 密码破解例1--- OrakelCrackert ...