Multiple Database Block Sizes and the Buffer Cache
In oracle 10g we can have multiple block sizes at the same time. When a tablespace is created we can assign a block size for the objects that will be created in that tablespace.
The DB_BLOCK_SIZE parameter in your initialization parameter file determines the size of your standard block size in the database and frequently is the only block size for the entire database.
The DB_CACHE_SIZE parameter in your initialization parameter file specifies the size (in bytes) of the cache of the standard block sized buffers. Notice that you don’t set the number of database buffers; rather, you specify the size of the buffer cache itself in the DB_CACHE_SIZE parameter.
You can have up to five different database block sizes in your databases. That is, you can create your tablespaces with any one of the five allowable database block sizes.
But before you use non standard block size, you have to define the cache size for these non standard block size. We have a paramter called DB_nK_CACHE_SIZE for setting the cache size for non standard block size.
The new init.ora parameters that allow you to use non-default block sizes are:
DB_2K_CACHE_SIZE
DB_4K_CACHE_SIZE
DB_8K_CACHE_SIZE
DB_16K_CACHE_SIZE
DB_32K_CACHE_SIZE
Another classification for buffer cache is depending on the algorithm used to keep the contents into the cache. We have basically 3 types in this catagory.
1) DB_KEEP_CACHE_SIZE
2) DB_RECYCLE_CACHE_SIZE
3) DB_CACHE_SIZE
DB_KEEP_CACHE_SIZE is where the object are always present when they are loaded. The objects which qualifies for this cache are those which are very frquently accessed and which has to be retained in memory. For example, frquently used small lookup tables. This cache is a subset of default cache defined by parameter DB_CACHE_SIZE. For any database we need to have DB_CACHE_SIZE set.
DB_RECYCLE_CACHE_SIZE is where you dont want to store the object. You want to clear off the object from cache as soon as it is used. You have to be careful while using this, since this may incure performance hit in case you allocated a frequently used object to this cache.
DB_CACHE_SIZE is the size for default cache.
it is important to note that the init.ora parameters and functionality regarding the keep and recycle buffer pools has changed between Oracle8i and Oracle9i. Those changes are
summarized in the table below:
Version Of Oracle Init.ora parameters Functionality
Oracle8i BUFFER_POOL_KEEP = <buffers> Subsets of the data
BUFFER_POOL_RECYCLE = <buffers> buffer cache
Oracle9i and 10g DB_KEEP_CACHE_SIZE = <size> Independent of the
DB_RECYCLE_CACHE_SIZE = <size> data buffer cache
To specify the use of the keep, recycle or default buffer pools, you can use the storage clause of the alter table statement:
alter table <table_name> storage (buffer pool keep);
alter table <table_name> storage (buffer pool recycle);
alter table <table_name> storage (buffer pool default);
Note: The keep and recycle buffer pools are only available for the standard
block size. Non-standard block-size caches have a single default pool.
So again back to non standard cache size. Lets say the default block size is 8K and you want to create 1 more block size for you future tablespaces. In that case you have to
assign the buffer cache for those block size in the memory. Remember, when
you create a non standard block sizes, the memory (cache size) allocation for these block
size will be taken again from physical memory RAM and hence the RAM consumption with go up. This memory wont be allocated from existing db_cache_size.
Example
SQL> show parameters db_cache_size
NAME TYPE VALUE
———————————— ———– ——————————
db_cache_size big integer 200M
Now we have 200M set for DB_CACHE_SIZE and db_keep_cache_size and db_recycle_cache_size is not set.
SQL> show parameter db_keep_cache_size
NAME TYPE VALUE
———————————— ———– ——————————
db_keep_cache_size big integer 0
SQL> show parameter db_recycle_cache_size
NAME TYPE VALUE
———————————— ———– ——————————
db_recycle_cache_size big integer 0
Also we can see the size of buffer cache using show sga
SQL> show sga
Total System Global Area 1073741824 bytes
Fixed Size 1984184 bytes
Variable Size 750786888 bytes
Database Buffers 209715200 bytes
Redo Buffers 6397952 bytes
Lets now try to create a tablespace with 4K block size.
SQL> create tablespace test_tbs4k datafile ‘/dy/oracle/product/db10g/dbf/test_tbs4k_1.dbf’ size 100M blocksize 4K;
create tablespace test_tbs4k datafile ‘/dy/oracle/product/db10g/dbf/test_tbs4k_1.dbf’ size 100M blocksize 4K
*
ERROR at line 1:
ORA-29339: tablespace block size 4096 does not match configured block sizes
Oracle in not intelligent enough to translate the block size of 4k into 8k buffer size. Because it cannot load the content of this datafile created for this tablespace having block size of 4k into buffer buffer of 8k.
So we need to create a buffer of 4K block size, after that only we can create a tablespace for 4k block size.
If we see the parameter db_4k_cache_size is not set.
SQL> show parameters db_4k_cache_size
NAME TYPE VALUE
———————————— ———– ——————————
db_4k_cache_size big integer 0
SQL> alter system set db_4k_cache_size = 100M;
System altered.
SQL> show parameter db_4k_cache_size;
NAME TYPE VALUE
———————————— ———– ——————————
db_4k_cache_size big integer 100M
SQL> show parameter db_cache_size
NAME TYPE VALUE
———————————— ———– ——————————
db_cache_size big integer 200M
SQL> show sga
Total System Global Area 1073741824 bytes
Fixed Size 1984184 bytes
Variable Size 750786888 bytes
Database Buffers 314572800 bytes
Redo Buffers 6397952 bytes
SQL>
If we see the above stats, it clearly shows tghat db_cache_size has not reduced, but
database buffer size has increased not to 300M. Now we can create a tablespace for 4K block size
SQL> create tablespace test_tbs4k datafile ‘/dy/oracle/product/db10g/dbf/test_tbs4k_1.dbf’ size 100M blocksize 4K;
Tablespace created.
SQL>
You can get the information about your database buffer cache from a view v$buffer_pool
SQL> select name, block_size, current_size from v$buffer_pool;
NAME BLOCK_SIZE CURRENT_SIZE
——————– ———- ————
DEFAULT 8192 200
DEFAULT 4096 100
You can get more stats and information on you buffer pool using the view v$buffer_pool_statistics.
Multiple Database Block Sizes and the Buffer Cache的更多相关文章
- Oracle优化 -- 关于Database Buffer Cache相关参数DB_CACHE_SIZE的优化设置
select size_for_estimate, buffers_for_estimate ,ESTD_PHYSICAL_READ_factor,ESTD_PHYSICAL_READS from v ...
- BUFFER CACHE之主要的等待事件
原因:资源紧张,等待其释放. 原因的原因:1. lgwr和DBWn进程写太慢:2. Buffer和latch不可用 原因的原因的原因:全表扫描.library cache latches数太多等. 视 ...
- Tuning 04 Sizing the Buffer Cache
Buffer Cache 特性 The buffer cache holds copies of the data blocks from the data files. Because the bu ...
- 相克军_Oracle体系_随堂笔记005-Database buffer cache
本章主要阐述SGA中的Database buffer cache. Buffer cache { 1.pin (读写瞬间的状态) 2.clean 3.dirty 4.unused } --Databa ...
- BUFFER CACHE之调整buffer cache的大小
Buffer Cache存放真正数据的缓冲区,shared Pool里面存放的是sql指令(LC中一次编译,多次运行,加快处理性能,cache hit ratio要高),而buffer cache里面 ...
- 查看buffer cache命中率
SQL> select name,value from v$sysstat where name in('db block gets','consistent gets','physical r ...
- Buffer cache 的调整与优化
Buffer cache 的调整与优化 -============================== -- Buffer cache 的调整与优化(一) --==================== ...
- Buffer Cache 原理
在将数据块读入到SGA中,他们的缓冲区被放置在悬挂散列存储桶的链表中(散列链),这种内存结构由大量 子cache buffers chains锁存器(也称为散列锁存器或CBC锁存器)保护. Buffe ...
- Memory Architecture-SGA-Database Buffer Cache
启动instance:1.分配内存空间SGA 2.启动后台进程 内存结构:1.SGA 2.PGA 3.UGA 4.Software code areas SGA components:1.Databa ...
随机推荐
- Java之集合(十六)ArrayBlockingQueue
转载请注明源出处:http://www.cnblogs.com/lighten/p/7427763.html 1.前言 JDK5是一个重要的更新版本,其提供了大量的并发类.之前的介绍都是一些util下 ...
- Spring Security构建Rest服务-0900-rememberMe记住我
Spring security记住我基本原理: 登录的时候,请求发送给过滤器UsernamePasswordAuthenticationFilter,当该过滤器认证成功后,会调用RememberMeS ...
- python基础笔记之面向对象
# class Foo:# name="kevin"## def __init__(self,puppy):# self.tomato= 'red'# self.dog = pup ...
- 【树】Binary Tree Zigzag Level Order Traversal
题目: Given a binary tree, return the zigzag level order traversal of its nodes' values. (ie, from lef ...
- start with connect by prior
start with connect by prior的使用: select … from tablename start with 条件1 connect by 条件2 where 条件3; sta ...
- Spark编程环境搭建(基于Intellij IDEA的Ultimate版本)(包含Java和Scala版的WordCount)(博主强烈推荐)
福利 => 每天都推送 欢迎大家,关注微信扫码并加入我的4个微信公众号: 大数据躺过的坑 Java从入门到架构师 人工智能躺过的坑 Java全栈大联盟 ...
- Spring Boot 的彩色日志
springboot的彩色日志灰常漂亮, 看起来也很舒服. 但是自定义的日志就是一纯白色的, 丑到不行. 所以就copy他的彩色日志来养眼: <!-- 彩色日志 --> <!-- 彩 ...
- lucene源码分析(2)读取过程实例
1.官方提供的代码demo Analyzer analyzer = new StandardAnalyzer(); // Store the index in memory: Directory di ...
- JAVA-7NIO之Socket/ServerSocket Channel
一.ServerSocketChannel Java NIO中的 ServerSocketChannel 是一个可以监听新进来的TCP连接的通道, 就像标准IO中的ServerSocket一样.Ser ...
- Kafka安装kafka-manager
1 .下载kafka-manager 想要查看和管理Kafka,完全使用命令并不方便,我们可以使用雅虎开源的Kafka-manager,GitHub地址如下: https://github.com/y ...