Hello顾问

'Six years ago we made the decision to re-write our ERP and CRM suit for the cloud. We called it Fusion. SAP called it confusion,' Ellison said. 'It will take years for SAP to catch up.'

Instead, Ellison said, SAP spent its energies building HANA, an in-memory database that competes with Oracle.

'When SAP, and, specifically Hasso Plattner, said they're going to build this in-memory database and compete with Oracle, I said. God, get me the name of that pharmacist, they must be on drugs. That was interpreted by Hasso as Larry doesn't believe in in-memory databases. Quite the contrary. We've been working on in-memory databases for 10 years. We have the world's leading in-memory database. It's called TimesTen.'

Ellison said.

'Six years ago we made the decision to re-write our ERP and CRM suit for the cloud. We called it Fusion. SAP called it confusion,' Ellison said. 'It will take years for SAP to catch up.'

Instead, Ellison said, SAP spent its energies building HANA, an in-memory database that competes with Oracle.

'When SAP, and, specifically Hasso Plattner, said they're going to build this in-memory database and compete with Oracle, I said. God, get me the name of that pharmacist, they must be on drugs. That was interpreted by Hasso as Larry doesn't believe in in-memory databases. Quite the contrary. We've been working on in-memory databases for 10 years. We have the world's leading in-memory database. It's called TimesTen.'

SAP HANA的白皮书中这样描述:
Column store is recommended if:
•• Calculations are executed on a single column or a few columns
only.
•• The table is searched based on the values of a few columns.
•• The table has a large number of columns.
•• The table has a large number of rows, and columnar operations
are required (aggregate, scan, and so on).
•• The majority of columns contain only a few distinct values
(compared to the number of rows), resulting in higher compression
rates.
从以上可以看到SAP HANA对Column Store的要求相对比较苛刻，因此它的对行压缩应该是有限的支持，因为如果要使用行压缩那就要挑选满足Column Store条件的表，这样是自相矛盾的。对比Exadata的列压缩技术以及混合列压缩技术，SAP HANA的列压缩技术限制很大。

SuperCluster提供半配和满配两种选择，当然可以集连扩展。

半配2台数据库服务器，全配4台数据库服务器。
每台数据库服务器配置：
Each SPARC T4-4 compute node configured with:
 4 x eight-core SPARC T4 Processors (3.0GHz)
 1 TB Memory
 6 x 600 GB 10,000 RPM SAS Disks
 2 x 300 GB solid state disks
 4 x dual-port InfiniBand QDR
 4 x dual-port 10 Gb Ethernet

Storage:
半配3台Storage Server X2-2，全配6台Storage Server X2-2。

还配有一台Sun ZFS Storage 7320 Appliance:
7320 Dual Controller, each with:
 2 x four-core 2.4GHz Intel® Xeon® Processors
 24 GB Memory
 1 x dual-port InfiniBand HCA
 2 x 500 GB SATA HDD
 4 x 512 GB read optimized SSD
7320 Disk Shelf with the following:
 20 x 2 TB using high capacity SAS-2 7200 RPM disks
 4 x 18 GB write-optim

湖南150多个免费WiFi热点开通
By 潇湘晨报 2012.03.14 10:36
作为第十一届读者节十大主题活动之一，潇湘晨报联合湖南电信推出的“你好，城市”“数字湖南 无线城市”免费WiFi热点建设活动，覆盖范围逐步扩大。今明两天，全省14市州150多个免费WiFi热点将全面开通。

在这些地方，只要你选择连接湖南电信WiFi热点“ChinaNet”，在认证页面中输入账号xxcb，密码96360，或者账号chinanet，密码10000，就可以享受免费的WiFi无线宽带畅游网络了。

这才只是开始，无线城市的体验范围将逐步扩大，面向全省。目标是一年内在全省建设一万个免费WiFi热点。据了解，今明两天，包括长株潭在内的全省14个市州，将全面开通免费WiFi热点（娄底、郴州明日正式开通，其他12市州免费WiFi热点今日可免费使用）。

昨天下午，在长沙市红星国际会展中心，刘鑫打开笔记本，很快搜索到了湖南电信的WiFi热点信号“ChinaNet”，连上热点后在认证页面中输入账号xxcb和密码96360随即进入了上网页面。可以浏览网站、下载软件。“速度不错！”刘鑫打开QQ和一个朋友视频，轻松地聊起了天，“你知道我在哪儿吗？我

RDS is from SilverStorm (acquired by Qlogic), Oracle RAC places upon interconnect (via libskgxp) for DML lock traffic and, to a lesser degree, for parallel query data shipping.
iDB is based on libskgxp

No 1/4 and 1/2 rack ONLY full rack, can be up to 8 racks!

2 Database Servers with:
  8 x Ten-Core Interl Xeon E7-8870 Processors (2.4GHz)
  2 TB Memory
  Disk Controller HBA with 512MB Battery Backed write Cache
  8 x 300GB 10,000 RPM SAS Disks
  8 x InfiniBand QDR (40Gb/s) ports
  8 x 10Gb Ethernet Port Bases on the Intel 82599 10GbE Controller
  8 x 1 Gb Ethernet Ports
  1 x ILOM Ethernet Port
  4 x Redundant Hot-Swappable Power Supplies
DISK Capacity:
  High Performance: 100TB raw Disk data
                    Data Load Rate: Up to 12TB/Hour
                    Up to 50,000 Database Disk IOPS
          

SAP on Oracle Database统计信息收集不能使用Oracle数据库自带的DBMS_STATS包来进行，需要使用BR*Tools来执行，因为SAP有自己的Policy对不同的数据量的表、索引使用不同类型的收集模式：

The update statistics method to be used the next time update statistics runs:

• E
  Estimate table statistics including indexes
• EH
  Estimate table statistics and create histograms
• EI
  Estimate table statistics and verify index structure
• EX
  Estimate table statistics, create histograms, and verify index structure
• C
  Calculate table statistics including indexes
• CH
  Calculate table statistics and create histograms
• CI
  Calculate t

/nSMLT
language management
overview sap support language

/nST04
database performance analysis overview

=======for User
/nsu01
create user id.

/nPFCG
create role
what is profile?

/nSU53
check the unsuccessful privilege of last login

/nSUIM
user information system


========for system analyze
/nST01
trace system internal activity

/nST03
workload in system

/nST06
operationg system monitor

/nRZ10
maintain the parameter file
instance profile
default profile
startup profile

/nSM36
define job

/nSE01
transport manage

/nSSAA
system manage assistant. it shows all the TCODE of activity about system.

/nSA38
the interface of clear the log

/nSM02
send a message to all user

最近一客户希望实现如下需求:

源端USER_A和USER_B同时对TABLE_A有全部的操作，客户希望不要复制USER_B对TABLE_A的DELETE操作，其他操作都需要复制：

方法1：使用FILTER

 

TABLE csd.TABLE_A,

FILTER (ON DELETE, @IF (@STREQ (@GETENV ('TRANSACTION', 'USERNAME'), 'USER_B'), 0, 1));

方法2：使用TRANLOGOPTIONS，但方法2需要对业务作限制，限制USER_B只能作DELETE的操作。

TRANLOGOPTIONS EXCLUDEUSER <user name>