2008-01-14 Mon
一、 参考资源,配置
FTP搜索引擎的安装配置(1)
FTP搜索引擎的安装配置(2)
就是与程序包里README.cn文件的内容一样的.
parker程序包下载
配置的过程参考这篇文章就可以了,启动了apache服务与ftp服务就可以; 遇到一个故障,修改Makefile文件里一行代码后顺利完成安装(见下文问题与解决);
搜索引擎dump 服务器上的目录结构提供搜索;用生成ls-lR文件传递给搜索引擎方式可以个性化设置搜索(通过裁剪ls-lR文本的内容来实现);
二、 安装涉及到的资源
目录:
/var/parker —-程序包目录
/var/www/cgi-bin/parker —-网页根目录
配置文件:
/usr/local/apache2/conf/httpd.conf —-apache配置文件
/etc/vsftpd/vsftpd.conf —-vsftpd配置文件
三、 启动服务
需要用到的服务是httpd , vsftpd,是RHEL系统默认都有安装了的;
打开服务的开机启动,启动服务
Chkconfig httpd on
Checkconfig vsftpd on
Service httpd start
Service vsftpd start
四、 VSFTP配置
Vsftp是系统默认自带的,进程名是vsftpd,需要修改如下文件
vsftpd.conf文件
#只允许admin连ftp,禁止了其他用户
anonymous_enable=NO
local_enable=YES
write_enable=YES
local_umask=022
dirmessage_enable=YES
xferlog_enable=YES
connect_from_port_20=YES
xferlog_std_format=YES
pam_service_name=vsftpd
listen=YES
tcp_wrappers=YES
#文件/etc/vsftpd.user_list 里的用户都不能访问ftp
userlist_deny=YES
#本地所有帐户都只能在自家目录
chroot_local_user=YESanon_root=/disk1/dbfile
anon_upload_enable=NO
anon_mkdir_write_enable=NO
anon_other_write_enable=NO
anon_world_readable_only=YES
local_max_rate=500000
anon_max_rate=300000
idle_session_timeout=3600
data_connection_timeout=1200
accept_timeout=60
connect_timeout=60
五、 遇到的问题与解决
1) Make install 语法错误
在src 目录下编译的时候提示语法错误
[root@dbbak src]# make install
cp search /var/www/cgi-bin/parker
cp statist /var/www/cgi-bin/parker
cp submit /var/www/cgi-bin/parker
cp siteinfo /var/www/cgi-bin/parker
/bin/sh: -c: line 2: syntax error: unexpected end of file
make: *** [install] Error 2Makefile 文件里原:
@if [ ! -d ../bin ]; then
mkdir -p ../bin; \
fi
修改为一行:
@if [ ! -d ../bin ]; then mkdir -p ../bin; fi
2) IE 浏览时候UTF8代码乱码
修改apache配置文件httpd.conf
修改默认字符为 GB2312:
#AddDefaultCharset UTF-8
AddDefaultCharset GB2312
3) 图片无法显示与网页链接错误
修改apache配置文件httpd.conf, 文档路径两处:
DocumentRoot "/var/www/cgi-bin/parker"
<Directory "/var/www/cgi-bin/parker">
Allow from all
</Directory>
4) 修改网页文字(有些文字需要修改源码)
Language.h 文件里修改,”Parker FTP搜索引擎” 为 “返回搜索首页”
5) dump目录结构的命令(flashdata)
方式一:
collect –r 直接dump目录结构
[root@dbbak bin]# cat flashdata
#!/bin/sh
/var/parker/bin/collect -r
/var/parker/bin/makestat
方式二:
collect –dr 获取ftp上的ls-lR文件来构建搜索内容; 用ls命令生成文本文件 ls-lR,可以过滤不需要被搜索到的文件
[root@dbbak bin]# cat flashdata
#!/bin/sh
/var/parker/bin/collect -dr
/var/parker/bin/makestat
首批新书在年会上与读者见面,感谢出版社与编辑,出版质量很不错;感谢IT168与ITPUB,提供了一个让大家聚在一起交流的机会;感谢各位读者,感谢他们对这本书的支持;感谢一起过去的各位同事,不辞辛苦的帮我搬书;感谢……
新书现在可以从如下地方网上购买:
新书宣传与售书活动
eygle接收新书
进行dml时,如果有任何违反约束的记录出现,整个dml操作会终止并开始回滚,这对于大批量的数据处理,不管是从时间还是系统性能消耗上来说,都是相当大的代价。我曾经有过因为几条数据主键冲突而导致几千万的数据导入失败的经历,通常这很难让人接受。为此,10.2版本中引入DML error logging特性,通过记录dml运行过程中出现的异常而不回滚事务,通过延迟对异常数据的处理来保证整个事务的顺利运行。
DML Error Logging 适用在INSERT, UPDATE, MERGE, and DELETE 语句。
1.创建Error Logging错误日志表
SQL> EXECUTE DBMS_ERRLOG.CREATE_ERROR_LOG(’dml_test’, ‘dml_test_error’);
2.Dml中加入Error logging子句
reject limit用来限制dml的最大报错数,若报错超出了最大限制数,dml语句中止并回滚,默认值是0,表示dml运行中遇到任何错误就终止回滚,你可以指定为unlimited。在并行dml中,reject limit设置为每个进程的最大报错数。
SQL> insert into dml_test select t.user_id, t.username from dba_users t log ERRORS INTO dml_test_error REJECT LIMIT unlimited;
例子:
Connected as TEST
--创建Error Logging表
SQL> EXECUTE DBMS_ERRLOG.CREATE_ERROR_LOG('dml_test', 'dml_test_error');
Table created
--主键冲突,整个Insert回滚
SQL> insert into dml_test select t.user_id, t.username from dba_users t;
ORA-00001: unique constraint (SYS.PK_DML_TEST_ID) violated
--利用DML Error Logging特性,导入成功
SQL> insert into dml_test select t.user_id, t.username from dba_users t
2 log ERRORS INTO dml_test_error REJECT LIMIT unlimited;
10 rows inserted
--查看异常数据记录表
SQL> select ora_err_number$ num,substr(ora_err_mesg$,1,40) as err_msg, user_id,username from dml_test_error;
NUM ERR_MSG USER_ID USERNAME
- ---------------------------------------------------------
1 ORA-00001: unique constraint(SYS.PK_DML 40 ORATEXT
1 ORA-00001: unique constraint(SYS.PK_DML 39 CTXSYS
.....
SQL>
--Insert成功后,通过dml_test_error来查询,处理失败的记录。
这个功能对于大批量数据Insert,Update还是很有用的。当然,也不是所有异常都可以通过DML Error Logging来处理,10.2版本只支持下列情况:
1.行值过长(Column values that are too large)
2.约束冲突(NOT NULL, unique, referential, and check constraints)
3.触发器执行过程中出现的错误(Errors raised during trigger execution)
4.Partition mapping errors
5.Certain MERGE operation errors (ORA-30926: Unable to get a stable set of rows for MERGE operation.)
这两天从Google Reader的Friends’ Shared Items中,好几个朋友都分享了同一篇文章,题目是”RoR部署方案深度剖析“,因为自己对这个内容也很感兴趣,仔细读了一下.应该说是很有意思的一篇文章,我也从中学到了不少东西,但是也发现了一些问题:
Apache/Nginx的接收缓冲区都只开了8KB,如果页面比较大,Mongrel就没有办法一次性把数据全部推给Web服务器,必须等到Web服务器把接收缓冲区的8K数据推到客户浏览器端以后,清空缓冲区,才能接收下一个8KB的数据。这种情况下,Mongrel必须和Web服务器之间进行多次数据传输,才能完成整个Web响应的过程,显然没有一次性把页面数据全部推给Web服务器快。
只有几个问题,application layer的读缓存和读的次数是否影响到socket另一端写socket的次数的? 写socket的次数是否和另一段socket read的次数对应? Web服务器向到客户端socket的写操作在什么情况下会被阻塞? 文中的一些观点和说法好像多少混淆了应用层和传输层的概念.
我们假设使用服务器端程序控制带权限的文件下载,某用户下载的是一个100MB的文件,该用户使用了多线程下载工具,他开了10个线程并发下载,那么每个线程Mongrel在响应之后,都会把整个文件读入到内存的StringIO对象当中,所以总共会创建出来10个StringIO对象保存10份文件内容,所以Mongrel的内存会一下暴涨到1GB以上。而且最可怕的是,即使当用户下载结束以后,Mongrel的内存都不会迅速回落,而是一直保持如此高的内存占用,这是因为Ruby的GC机制不好,不能够及时进行垃圾回收。
Mongrel处理rails request和静态文件使用的是不同的handler.处理静态文件用的是DirHandler,DirHandler没有通过调用response的start方法使用到StringIO,而是直接调用了response的send_file方法,直接write socket.验证过程非常简单,自己起一个mongrel,在public下放一个100M的文件,通过浏览器下载一下.在我的mac上,本来是76M的mongrel,在下载结束后是74M,下载过程中也一直小于80M,最低到过68M.无论是过程中还是结束后,都没有出现占用内存暴涨的现象.
另外,mongrel本不应该用来处理静态文件,该文章开始的图也画得很清楚.
我们知道Mongrel在使用Lighttpd的时候,可以达到最快的RoR执行速度,但是Lighttpd当前的1.4.18版本的HTTP Proxy的负载均衡和故障切换功能有一些bug,因此一般很少有人会使用这种方式。
Lighttpd 1.5重写了mod_proxy_core,sqf方式下的balancer已经没有这个问题.
而Mongrel就简单了,gem install mongrel安装完毕,mongrel_rails start启动,哪个人不会?毕竟绝大多数开发人员和部署人员不是高手,他们熟悉哪种方式,自然就会推崇哪种方式。
首先,我不太赞同复杂的就好,简单的就不专业这样的观点,Unix philosophy也不是这样的.其次,mongrel也不是mongrel_rails start就能部署到产品环境的.
2008年1月14日,存储部落所在的bluehost主机被黑客入侵,窃取了大量的资料后,所有重要的数据惨遭删除,使得存储部落从下午13:30分左右停止运行,知道晚上23时左右才恢复,停止运行时间长达9个半小时。
在此向这个无耻的黑客及其无耻的行为表示强烈的谴责!
令人万分高兴的是bluehost每天都会进行一次数据备份,所以可以通过数据操作将数据rollback到13日13时左右。对于存储部落来讲,所有程序文件、图片和文档都没有丢失,数据库丢失了十几条记录(6位朋友的注册记录和6条留言记录),总体损失不大,只要要麻烦这6个朋友费神在注册一次了。
从此可以看出,重要的数据系统除了要具备良好地防病毒、防黑客攻击措施外,还要建立良好数据安全备份系统。有了数据备份系统,即使数据丢失了,也能够快速地将数据恢复回来。可惜bluehost的备份周期是24小时,假如备份周期是12小时、2小时、或者2分钟,那可能一点数据都不会丢失了。
呵呵,所以备份系统真的很重要啊!
作者:Fenng 发布在 dbanotes.net. 订阅 DBA notes
早晨回家,天还蒙蒙亮。走到东方威尼斯附近的时候,忽然发现卖煎饼的夫妻俩已经准备好摊子了。今天有点冷,夫妻俩偎依在那里,看着好温馨。旁边的东方威尼斯也算是杭州的著名消费场所了,即使是清晨,也停了不少不错的汽车。
每天早晨上班的时候,都看到好多人在那里排队买煎饼,自己因为痛风不能多吃"油炸"性质的东西,也不愿意等太久,还真没吃过几次他们的东西。Laura 上班走的比我早,经常买来吃。告诉我好几件有趣的事情。
夫妻俩一人一个摊子,有的时候人多,哪些偶尔路过的人就会随机选一个摊子。可能以为这夫妻俩是竞争关系,所以"选择"后经常是有一队人特别多,后来的人还偏偏选人多的队伍等(估计是以为人多的那个煎饼好吃),这时候夫妻俩面前人多的那个就会小声的说:排另一个吧,我们是一家的。
有一段时间,连续几天都没看到他们的煎饼摊。后来总算出来了,Laura 趁他们不忙的时候问了一下,原来是被城管把车子收走了。“交了好几百才把车子要回来。",那个妻子叹口气说。
--EOF--
相关文章|Related Articles
评论数量(0)|Add Comments
本文网址:http://www.dbanotes.net/mylife/common_people_life.html
最近作者还说了什么? Follow Twitter / Fenng
In the second of a three-part series, Alan Cox talks about community and the enterprise. See the first episode.
Author:NinGoo posted on NinGoo.net
两天的2008IT技术精英年会热热闹闹的过去了,虽然一身疲惫的回到了杭州,却也收获颇丰。今年的年会无论是会务的组织还是演讲的水平,较之第一届都有不少的提升,感谢的话就不多说了,还是挑些毛病吧。
1.九华山庄离城里实在是太远了点,因为正好有几个9年未见的高中兄弟在京城出差,所以周六趁着这个机会聚了聚,2顿饭吃了近7个小时,可见孔夫子说有朋自远方来不亦乐乎是有道理的。而且到北京三次了,这次总算抽空逛了下小学一年级就开始念叨的天安门,而且正好碰上降旗仪式,也算是见识了国家礼仪。这么几下折腾,加上来回车程,回到宾馆已经冷冷清清,只好一个人洗洗睡了。
2.九华山庄没有像去年香山宾馆一样的大厅,参会的朋友们之间也就少了一个群聊的机会,甚为遗憾。所以在第二天上午主会场里,就出现了台上厂商慷慨陈词,台下三五成群相谈甚欢的局面。这样多少是不礼貌的,所以中途几次和几个朋友偷偷溜出会场叙旧,又被工作人员“请回”。
3.周日中午的菜,那是相当的合我的胃口,不过估计不吃辣的朋友要饿肚子了^_^
4.这里要严重向老虎抗议,三星打印机的奖品实在太重了,加上piner的新书,去时只背了一个空袋子,回来时却肩抗手提的三大包,可累坏了,呵呵
说到piner的新书,等了这么长时间,终于拿到了亲笔签名版了,翻了一下,书的印刷质量相当的不错,内容自不必说,丰富而实用,大多是piner这几年来实际工作经验的总结,对于想在oracle方面深入发展,想了解从存储到主机到oracle整个体系架构的高可用性方面技术的朋友们来说,这本煌煌80万言的巨著,真的是一份难得的礼物。而且piner对于写作的精益求精的态度,也让我深感佩服,三个多月前我来到淘宝的时候,书已经算是基本完稿,但是这三个月来,直至到付印前夕,piner都还是在坚持不停的修正,我刚看了看以前已经看过的几章,内容都有了不少的补充。
不过这本书,很多内容都是高度概括的,并且从体系架构的角度来看问题,所有对于初学者来说不容易看明白,必须需要有足够的概念和基础知识。不过没有关系,好的书都需要反复的咀嚼消化,不但需要记忆和理解,更需要思考与质疑,如果在看书的过程中有什么疑问,不用犹豫,直接去piner的个人网站www.ixdba.com留言,当然也可以去我们团队blog一起交流心得。
PS:发现书中一处小错误,ora-04031都印成ora-0431了,呵呵。
Related Articles
ITPUB年会暨2008中国IT技术精英年会已经结束,两天的会议下来,匆匆聚散,点点滴滴萦绕心头,聊记一二以为纪念。
1.温泉氤氲暖冬日
虽然是在九华山庄开会,但是泡温泉就只能在晚上了,12号那天是北京降温的开始,-12度左右的晚上,大家一起去泡温泉。
在更衣室里换衣服的时候已经颇为踌躇,看着天寒地冻的,披着浴袍一群人冲入温泉池中,实在是需要勇气,还好等自己进入温泉池里,发现并没有想像中那么冷,不过我的确怀疑冬天泡温泉不是一个好主意。
温泉水暖起来的时候,水汽氤氲升起,如梦如幻,倒也别有一番景象,九华晚上又放烟花,在寒冷的天空里绽放绚烂,让人想起安妮宝贝写过的她比烟花更寂寞,那是一部电影的名字,英文名是《HILARY AND JACKIE 》,被翻译为她比烟花寂寞(另一个译名是狂恋大提琴),看过那部片子,的确深含寂寞之意。
将烟花与寂寞的意象联系在一起,不知是何人的杰作?
D.C.B.A和我和Julia企图去泳池里游泳,结果因为没带泳帽而被管理员赶了出来,我心中窃喜,还好没让我丢脸。D.C.B.A游了10米就惊叹体力不支,记得我以前说过健康比什么都重要,大家还是要常记心头的好:)
四个小时的温泉限时,不到一个小时我们就逃之夭夭,就只是泡着也受不了4个小时的蒸腾,我们选择了快速逃跑。
2.新朋旧友话来年
温泉出来,赶快回房间补充水分,楼方鑫同学和他的同事范向荣(eagle_fan)来我房间聊天,eagle_fan是一个比我们小得多的年轻人,但是颇受诸多前辈的激赏,是年轻DBA中的出色人物。
老楼戏言,以后我们也不用研究技术了,有问题就问这些年轻人好了:)
我和老楼聊得最多的是2008年的个人计划,对于我们这一代的DBA,时间和年龄会让我们迅速老去,所以如何规划自己的未来就显得更加迫切。
在2008年,我个人会有一些新的计划,老楼仗义助拳,声称会全力支持我,嘿嘿,我的底气足了不少。
。。。未完待续。。。
相关文章|Related Articles
评论数量(3)|Add Comments
本文网址:http://www.eygle.com/archives/2008/01/annual_conference_mem_one.html
Coghead, like DabbleDB and Zoho Creator, allows you to quickly create a hosted database-backed web application without programming. But unlike DabbleDB and Zoho Creator, Coghead wants to create a platform not just for web applications but for web businesses, and they’re doing it on top of Amazon Web Services (AWS).
Coghead is building their business on a model similar to that of AWS: providing agile and scalable services over the web using pay-as-you-go pricing. Through their affiliate program, Coghead makes it easy for more software-as-a-service (SaaS) web businesses to launch, just like Amazon wants to use their limited beta DevPay service to encourage SaaS businesses atop AWS.
In Coghead on AWS, you can see the arrival of the SaaS ecosystem, as web-based services combined and layered on top of each other make it easier and cheaper than ever for anyone with a good web business idea to turn it into a revenue stream.
Coghead unveils v2.0 on Amazon Web Services
Today Coghead unveils version 2.0 of its do-it-yourself web application platform. Their 50 new features and enhancements look less interesting than their migration to Amazon Web Services for their computing infrastructure. (CTO Greg Olsen, who sees a SaaS-based ecosystem as the inevitable future of computing, calls AWS the most significant product of his lifetime.)
To build an application with Coghead, you drag and drop widgets to make web forms via Coghead’s Flex-based user interface. You don’t have to do any programming because Coghead has done that for you, nor do you have to buy and manage your own hardware and hosting. Coghead handles that for you, too — by delegating such responsibilities to AWS and to services built on top of AWS.
Web businesses, not just web applications
In December, Coghead launched the Coghead Affiliate Program, which provides system integrators, value-added resellers, and web tycoon wannabes an easy way to build and sell web applications using a SaaS model. Coghead affiliates build applications using Coghead’s do-it-yourself interface, then set their own per-user, per-month prices for use of those applications. Coghead takes their part of the monthly user fees, while the application developer gets the rest as revenue.
Until now, there hasn’t been an easy, low-cost way for a tech-savvy person with a good niche web application idea to turn that into a business. You’d have to find programmers, arrange web hosting, and build a payment system into your application. Coghead does all that for you.
The next web expansion?
Coghead builds on the SaaS offerings of others, including AWS (which offers both hardware and software as a service), as well on other services built on top of AWS — such as offerings from RightScale. Coghead in turn creates an environment for yet more SaaS businesses to launch. Current affiliates include Hekamedia for education, allRounds for private equity, and MCF Technology Solutions for general business applications.
Coghead is just one of many companies growing in the SaaS ecosystem. I previously covered Workplace2Go, which offers hosted business applications and uses the Jamcracker service delivery network behind the scenes. Workplace2Go, like Coghead, reduces the friction of starting a business.
It’s too early to tell which players in the SaaS ecosystem will survive and thrive — but Coghead, as both a consumer and producer of service-based offerings, looks well-positioned for success in this cost-efficient new world.

有些事情,就是那样.
结束了,一切都结束了.
只想说,保重!
只是不知今夜我会醉倒在何方,不过这又有什么关系列?人生不胜一场醉,醉了,会忘记很多事情,不是吗?那么,就让我每日沉醉不要醒,这样,我就不会记得所有的快乐和悲伤.
Over the weekend I held a little competition here at Problogger where readers were asked to submit Twitter Style blog tips (tips that were 140 characters or less).
The winner (chosen randomly) is CatherineL who submitted this tip:
“Be human - Your readers want to learn about your mistakes, as well as your successes.”
Congratulations Catherine - I’ve just emailed you to get your address details.
Thanks to everyone who entered. There were over 200 tips submitted and among them were some real gems. I personally found the exercise to be a lot of fun to read through this afternoon - there’s some great tips in the mix.
If you have some spare time you might find reading through the comments worthwhile.
The Google Operating System blog has an interesting post on Google's scale based on an updated version of Google's paper about MapReduce.
The input data for some of the MapReduce jobs run in September 2007 was 403,152 TB (terabytes), the average number of machines allocated for a MapReduce job was 394, while the average completion time was 6 minutes and a half. The paper mentions that Google's indexing system processes more than 20 TB of raw data.
最近很多朋友抱怨我的blog上写的技术文章少了
其实我也发现这个问题,为什么少了?
并不是我懒,也不是没有内容可以写.实在是很多的技术问题,都直接和环境有关系,而客户的环境也是客户的商业机密之一.很多时候,我把握不好这个度,把握不好这个度,那就不如不写.
或许那天我能把握好这个度,或者能写些和客户环境没关系的技术文章,那个时候,我会多写点吧.
但是应该不是现在,因为现在太忙,根本没有时间去考虑这个度的问题.
因此请见谅!
众所周知,在安装Oracle Clusterware(Former Oracle CRS)之前,有一些必备的条件,比如双网卡,同版本的操作系统,一些必需的补丁等等,还有一些比如同样ID的组和用户,这些都可以通过clusterware附带的一个检查工具cluvfy进行检查。详细的列表和检查方法可以参考oracle的官方安装文档
除此之外,安装之前另外一个重要的前提就是ssh或者rsh的配置,这就是oracle在节点之间进行文件copy时候必须的,下面就分十个步骤介绍如何进行ssh的配置。
1. Login as oracle user
2. create .ssh directory in the oracle user’s home directory
$cd
$mkdir ~/.ssh
$chmod 700 ~/.ssh
3. Generate an RSA and DSA key for SSH
$/usr/bin/ssh-keygen -t rsa
$/usr/bin/ssh-keygen -t dsa
4. After this, four files generated
~/.ssh/id_rsa.pub
~/.ssh/id_rsa
~/.ssh/id_dsa.pub
~/.ssh/id_dsa
5. Create an authorized key file
$touch ~/.ssh/authorized_keys
$cd ~/.ssh
$ls
6. Copy the contents of rsa.pub and dsa.pub files to authrorized file
[oracle@node1 .ssh]$ ssh node1 cat /home/oracle/.ssh/id_rsa.pub >> authorized_keys
oracle@node1’s password:
[oracle@node1 .ssh]$ ssh node1 cat /home/oracle/.ssh/id_dsa.pub >> authorized_keys
[oracle@node1 .ssh$ ssh node2 cat /home/oracle/.ssh/id_rsa.pub >> authorized_keys
oracle@node2’s password:
[oracle@node1 .ssh$ ssh node2 cat /home/oracle/.ssh/id_dsa.pub >>authorized_keys
oracle@node2’s password:
7. Use scp to copy the authorized file to other nodes in ~/.ssh
[oracle@node1 .ssh]scp authorized_keys node2:/home/oracle/.ssh/
8. Change the permissions on the oracle user’s authorized file on all cluster nodes
$ chmod 600 ~/.ssh/authorized_keys
9. Enable OUI to use ssh and scp
$ exec /usr/bin/ssh-agent $SHELL
$ /usr/bin/ssh-add
10. Verify ssh configuration on all nodes
$ ssh nodename1 date
$ ssh nodename2 date
其他的具体步骤可以参考oracle的手册B28264-03,地址如下:
http://download.oracle.com/docs/cd/B28359_01/install.111/b28264/toc.htm
众所周知,在安装Oracle Clusterware(Former Oracle CRS)之前,有一些必备的条件,比如双网卡,同版本的操作系统,一些必需的补丁等等,还有一些比如同样ID的组和用户,这些都可以通过clusterware附带的一个检查工具cluvfy进行检查。详细的列表和检查方法可以参考oracle的官方安装文档
除此之外,安装之前另外一个重要的前提就是ssh或者rsh的配置,这就是oracle在节点之间进行文件copy时候必须的,下面就分十个步骤介绍如何进行ssh的配置。
1. Login as oracle user
2. create .ssh directory in the oracle user’s home directory
$cd
$mkdir ~/.ssh
$chmod 700 ~/.ssh
3. Generate an RSA and DSA key for SSH
$/usr/bin/ssh-keygen -t rsa
$/usr/bin/ssh-keygen -t dsa
4. After this, four files generated
~/.ssh/id_rsa.pub
~/.ssh/id_rsa
~/.ssh/id_dsa.pub
~/.ssh/id_dsa
5. Create an authorized key file
$touch ~/.ssh/authorized_keys
$cd ~/.ssh
$ls
6. Copy the contents of rsa.pub and dsa.pub files to authrorized file
[oracle@node1 .ssh]$ ssh node1 cat /home/oracle/.ssh/id_rsa.pub >> authorized_keys
oracle@node1’s password:
[oracle@node1 .ssh]$ ssh node1 cat /home/oracle/.ssh/id_dsa.pub >> authorized_keys
[oracle@node1 .ssh$ ssh node2 cat /home/oracle/.ssh/id_rsa.pub >> authorized_keys
oracle@node2’s password:
[oracle@node1 .ssh$ ssh node2 cat /home/oracle/.ssh/id_dsa.pub >>authorized_keys
oracle@node2’s password:
7. Use scp to copy the authorized file to other nodes in ~/.ssh
[oracle@node1 .ssh]scp authorized_keys node2:/home/oracle/.ssh/
8. Change the permissions on the oracle user’s authorized file on all cluster nodes
$ chmod 600 ~/.ssh/authorized_keys
9. Enable OUI to use ssh and scp
$ exec /usr/bin/ssh-agent $SHELL
$ /usr/bin/ssh-add
10. Verify ssh configuration on all nodes
$ ssh nodename1 date
$ ssh nodename2 date
其他的具体步骤可以参考oracle的手册B28264-03,地址如下:
http://download.oracle.com/docs/cd/B28359_01/install.111/b28264/toc.htm
Lifestreaming, according to Wordspy, is "an online record of a person's daily activities, either via direct video feed or via aggregating the person's online content such as blog posts, social network updates, and online photos." In this post we review some of the top lifestreaming web apps: Onaswarm, Lifestrea.ms, Soup, Jaiku (the service Google bought), and perhaps the most popular of them all, Tumblr.
There's even a niche blog devoted to lifestreaming, called The Lifestream Blog. It recently noted that Wired magazine named lifestreaming a "wired" technology (as opposed to 'tired' or 'expired'). So it seems lifestreaming is the new black. Let's check out some of the leading lifestreaming apps...
Tumblr
For a recent episode of Read/WriteTalk Sean Ammirati sat down with David Karp, the founder of Tumblr. Tumblr is a platform that makes it easy to create Tumblelogs - which Wikipedia defines as:
A variation of a blog, that favors short-form, mixed-media posts over the longer editorial posts frequently associated with blogging. Common post formats found on tumblelogs include links, photos, quotes, dialogues, and video. Unlike blogs, this format is frequently used to share the author's creations, discoveries, or experiences without providing a commentary. One of the many tumblelog sevices is tumblr.

Onaswarm
Onaswarm is a new lifestreaming application from Toronto's David Janes and BlogMatrix. Marshall Kirkpatrick wrote that Onaswarm is a smart, interesting service that combines groups, microformats and flashes of really good usability.
It's very text-centric and clearly better for geeks than it is for the artists who like Tumblr, for example. The Onaswarm site architecture and navigation need a substantial overhaul to improve usability, despite some nice touches.

Lifestrea.ms
Lifestrea.ms is a powerful new lifestreaming service from Germany that you'll want to keep an eye on. Marshall checked it out and said that it's a real testimony to the potential of the new web that anyone would even try to create something like this company has. Currently in private beta, we hope the company will fix its usability issues and launch soon. Send an email to beta@lifestrea.ms if you want on the list for an account.
Lifestreaming aggregates all your inbound and outbound activity online, see Tumblr or FriendFeed for other examples. If everything under the covers at Lifestrea.ms can be made as good as the front page of the site, then we'll be in great shape. That page alone is a marvel to witness.
Soup
Watch out Tumblr, here comes Soup. According to Josh Catone, Soup is an easy to use tumble blogging application that includes two killer features: social networking (kinda) and outside activity streams. It's sort of a cross between Tumblr, Pownce, and a social activity aggregator.
At its core, Soup is a microblogging app, and a pretty easy to use one. Their tumble blog set up supports text, link, quote, image, and video posts. Sign up is a snap (you can actually begin posting to your tumble blog before creating an account), and like Tumblr, Soup blogs can be mapped to an outside domain.
Jaiku
Jaiku can aggregate and automatically republish stories from your other activity streams: blog posts, del.icio.us links, Flicker photos, even Twitter updates. In this regard, it is a lot like Tumblr (another service that has a huge lead on it traffic-wise). We think this is the part of Jaiku that Google was interested in when it purchased the site -- Jaiku as an activity stream aggregator, not Jaiku as a presence app.

We heard last summer about a Google sponsored project at Carnegie Mellon University called "Socialstream." Socialstream's goal was to "create a system for users to seamlessly share, view, and respond to many types of social content across multiple network." The idea was basically for Socialstream to be a hub for all of your social networking activity -- whether that was on Facebook, MySpace, Twitter, Digg, or Flickr -- all of your attention data would be collected in one place where you could manage and share it.
Finally, you may want to check out What's Next on the Web: a RWW Toolkit for 2008, which features Open Data as one of the 5 big trends Marshall Kirkpatrick compiled resources for. Lifestreaming is a type of web app that will benefit greatly from open data, so check out our toolkit to prepare yourself.
作者:谢国忠 | 评论(5) | 标签:美国, 经济, 香港股市, 内地股市, 泡沫, andy, xie, 谢国忠
(译文在全文后)
The United States is sliding into recession. Japan seems too. China is tightening aggressively. Oil has soared past $100 per barrel. The stars are lined up for a perfect storm as 2008 begins. Hang Seng Index may drop 20% over the next three months. The market may recover in the second quarter as the Olympic mania affects market sentiment. Before investors can taste the joy of 2008, they must eat bitterness first.
2008 is a turning point in the global economy. The current bull market began in mid-2003: Americans recovered from the '9-11' shock and China recovered from the SARS shock. Both felt a near-death experience and became determined to live it up a bit, just in case. Americans went on a borrow-and-spend binge. The defining picture must be throngs of people rushing through the just-opened doors at Wal-Mart Centers. Chinese threw the money into property and then stock market and watched both rise. The defining picture must be the throngs of people staring up at the tickers screens. The two spent differently but got the same thrill.
Well, every party must end. America's is ending now. China's may have a few innings left but may experience a big hiccup soon as the US burst chills sentiment here and inflation scares Chinese policymakers into taking the away, even though temporarily. The combination may cook up the perfect storm for Hong Kong market.
The bull market depends on rising earnings and cheap capital. The former has benefited from 15% annual growth rate of China's nominal GDP for the past three years and rising share of corporate earnings in the economy due to faster asset appreciation. The macro tightening could take a big bite out of both. The tightening may slow nominal growth rate by 20%. It will reverse the rising trend of profit share in the economy, mainly due to its powerful effects on earnings in the property and financial sector. The growth rate for China's corporate earnings may halve in 2008, with banks and properties the biggest casualties.
Despite the Fed's rate reductions, the US economy is still sliding into a recession. The credit crisis has exposed the risks of buying into complex Wall Street products. It deters international capital from flowing to the US. Foreign capital has funded the US borrow-and-spend binge. Unless the confidence in Wall Street returns, Americans won't have enough money to spend. It takes time for foreigners to recover their faith in the Wall Street. The Fed's policy couldn't substitute this. Its rate reductions would only make dollar weaker and increase inflation. As a result, stagflation may stalk the US economy for a couple of years.
The US recession will decrease risk appetite among international investors. As international capital still dominates the Hong Kong market, it could see significant outflow. Further, many financial institutions in the US have capital shortage. They may have to pull money out of Hong Kong. Facing a banking crisis at home in 1998, Japanese financial institutions pulled big amounts of money out of Hong Kong. The US financial institutions may do the same in 2008.
Would Chinese money come to the rescue? Not soon enough. Chinese money may be flowing back into the A-share market recently. Even though the A-shares are so much more expensive than the same shares in Hong Kong, buoyant local sentiment still keeps them up. Chinese investors may be irrational. But, there are so many of them that they can keep an irrational market going for a long time. Hong Kong market is just too rational for the comfort of Chinese investors. Hence, Chinese market becomes a safe haven during the international correction.
The Fed reductions may not benefit like before. The property demand in Hong Kong depends on gains in the stock market rather than wage income. The wage gains are not exceeding inflation by much. The lower US interest rate won't spike Hong Kong wage earners to borrow and purchase properties like before. Stagnant population is another headwind for property. Japan shows that, with stagnant population, even zero interest rate doesn't make property price go up. Despite all the hypes about Hong Kong property, it may have peaked already.
In the second quarter, international financial markets may calm down, after fully pricing in a US recession. Chinese money may return to Hong Kong again, to play at a lower level. International investors may also have overcome their fear of the US dragging down China. As the two embrace, the market may fly again, even though the US market remains stone cold.
The Olympics party may not last long. After the summer, reality may catch up with investors again. Many Chinese companies are concept plays, not lasting franchises. Some might even be scams, despite being taken to the market by renowned investment banks. Remember subprime: big names don't mean much anymore. China's tightening will expose the negative cash flow businesses of such companies. The resulting explosions may affect confidence.
The air in Hong Kong and Shanghai may be squeezed soon after the Olypics. The deflating process may be painful to many. But it would make China's capital markets and economic development healthier. Many 'share gods' have emerged in Hong Kong and Shanghai. Their 15 minutes of fame are far more damaging than Paris Hilton's: they suck credulous housewives into overvalued shares without the pleasant look to soothe. Worse, they might be talking there own books and could be getting out while talking bullish.
Even bubble here goes through the same routine. Investment banks take doggy companies to the market for a fee. Some new faces suddenly don the cover of Fortune Magazine as the riches this and that. When the show stops, most stars suddenly vanish. Some manage to escape into the deep woods of Thailand with some stolen cash. Little people always lose. We seem to be doomed to repeat the same mistake again and again.
翻译:小编
美国已经开始了衰退期,日本看来也这样。中国也在很进取的紧缩。油价已经过了一百美元。当2008刚刚开始,闪烁的星星看来已经为一场完美的暴风雨做好了准备。恒生指数可能会在未来三个月下跌百分之二十。市场可能在奥运效应的刺激下,在第二季度出现复苏。但是在投资者尝到甜头之前,他们必须先吃到苦头。
2008是全球经济的拐点。现在的牛市是从2003年中开始的:美国从911,中国从非典的震惊中醒来。两者都经历了死亡的感觉,于是决定,不管怎样,需要好好活着。于是,美国开始了借贷消费,最典型的画面,就是刚刚开门的沃尔玛门口的长龙。而中国人则把钱投入了房地产,然后是股市,看着它们上升。典型的画面就是股价牌前的人龙。两者花钱的方式不同,但是却有相同的结局。
好了,晚会总有结束的时候。美国的已经结束了,中国的可能还有一点点时间,但是也可能很快会经历倒退。因为美国的影响,加上政策制定者对于通胀的恐惧,使得他们提前把风球挂上,即使只是暂时的。而这些的混合,可能会导致香港市场的暴风雨的到来。
牛市依靠增加的利润和便宜的资本。之前大家受益于中国过去三年超过百分之十五的经济增长,以及快速的资产折旧带来的利润分享。但是宏观调控会影响这一切。紧缩政策会让快速增长减慢大约百分之二十。这会让经济利润分享的上升趋势出现逆转,特别是在房地产以及金融业的影响。中国的企业盈利在2008年可能会减少一半,房地产和银行会成为最大的受害者。
尽管联储局减息,美国经济依然走向衰退。次案危机显示出了购买华尔街复杂金融产品的风险。这阻止了国际资本流向美国。而外资是美国的信贷消费的支撑。除非华尔街的信心恢复,不然美国没有足够的资金被消费。要恢复外资对于华尔街的信心需要时间。联储局的政策是没有用的。它的减息政策只会导致美元更加弱势以及通胀加大。而结果,美国会出现好几年的通缩。
美国的衰退会让国际投资者增加风险意识。由于国际资本仍然主导香港股市,因此会看到明显的流出。另外,美国的很多金融机构出现资金短缺,因此他们会从香港抽出资金。在1998年,日本银行因为缺乏资金,而从香港抽回大批资金,在2008年,美国的金融机构可能会这样做。
那末内地的资金会来拯救香港吗?可能时间不够快。中国内地的资金可能会在最近回到A股市场。尽管A股的价格远远比香港的H股要贵,但是乐观的市场气氛对他们很吸引。中国内地的投资者可能是非理性的。但是他们有足够的非理性的投资者来推进市场在非理性中上升。而香港市场对于那些中国内地投资者来说过于理性,而中国内地股市成为国际市场调整中的一个天堂。
联储局的减息的效果可能没有从前那样有效,香港房地产市场的需求,依靠的是股市而不是工资收入。工资的增长没有比通胀多多少。美国的减息并不能够像以前那样刺激香港的打工仔去购买房地产。人口老龄化也是房地产的影响因素,就好像日本,即使是零利率,也没有造成房价上升。对于乐观者来说,香港房价的顶峰可能已经抵达了。
到了第二季度,国际金融市场可能会冷静下来,在经历了美国衰退之后。内地的资金可能再回到香港,在底位进行投资。国际投资者可能已经克服了对于美国拖累中国经济的担心,香港股市可能复苏,但是同时,美国市场还是石头一块。
奥运效应可能不会持续长久。夏天之后,投资者会面对现实。很多中国的公司只是在玩概念,不是长久性的。很多可能只是一个空壳,除非真的被投资银行放入了市场。记得这样一句话:名声大不再意味着什末。中国的紧缩政策会让那些资金流不足的公司显现出来。而这些情况的暴露可能会影响信心。
香港和上海的空气在奥运后可能会紧张起来。通缩的过程可能会很痛苦。但是会让中国的资本市场发展得更加健康。很多“股神“已经淹没在香港和上海的股市中。他们的15分钟的操盘造成的损害远远要比PARIS HILTON严重的多:他们把那些家庭主妇的钱,吸引到了那些高估价的股票里面。更糟糕的事,在牛市的时候把这些钱占为己有。
泡沫总是用同样的方式发生。投资银行先把那些所谓的公司上市来赚回手续费。于是在一些财富杂志上就会出现一些新的面孔。当这样的表演停止,于是这些面孔就消失了。一些人就会逃到泰国的森林里面,带着不法得来的现金。小人物总是输的。我们必须避免同样的错误再次发生。
岩岩,今天又去远洋看你了,本来说要送你一捧大大的白色菊花,但是想了又想,觉得那是寄托哀思的方式,然而你现在是快乐的,所以,最后还是买了一包七星,自己点了一根,剩下的放在你房间的门口了,没帮你点,自己点哈。还送了你一张今天的《新京报》,毕竟,烟放在地下不太干净嘛。
在窗边站了很久,耳机里面的《JOIN ME》重复了不知道多少遍。从窗口往下照了一张照片,还真是高啊。
对了,还碰到了一个你的朋友,他也来给你送花了,我们没太多说话,只是握了握手,点了点头,说了几句,但是我知道我们还都有些不舍得你。
等你完整的走了吧,我和8P,还有姐姐说好了,要去一个漂亮的地方种一棵树,这是你的愿望 - “如果有来生,要做一棵树,站成永恒,没有悲欢的姿势。一半在尘土里安详,一半在风里飞扬,一半洒落阴凉,一半沐浴阳光。非常沉默非常骄傲 从不依靠从不寻找。”
如果有好多人都一起来种树呢,那么也许可以成就一片林子吧,这样的话,你就更不孤单了。
2008-01-13 Sun
2008-01-12 Sat
AnySQL.net
DBA notes
Oracle & Starcraft
eagle's home
给你点color see see
AnySQL.net English
Oracle Scratchpad
Oracle Life
OracleDBA Blog---我不在江湖,江湖却有我的传说!
Photos from dbanotes
Chanel [K]
xzh2000的博客
Oracle Security Blog
ERN空间
Eddie Awad's Blog
MySQL Performance Blog
The Tom Kyte Blog
del.icio.us/fenng/oracle
AIXpert
O'Reilly Databases
Red Hat Magazine
DBASupport
DB2 Magazine 中文版
developerWorks : AIX 专区的文章,教程
Pythian Group Blog » Log Buffer
车东[Blog^2]
blue_prince
玉面飞龙的BLOG
此生 今世
人生就是如此
Orange Tiger 木匠 的 移民生活
生活帮-LifeBang
Hey!! Sky!
dba on unix
Oracle Notes Wiki
Welcome to brotherxiao's Home
柔嘉维则@life.oracle.eng
Fenng's shared items in Google Reader
jametong's shared items in Google Reader
缥缈游侠-logzgh
Tanel Poder's blog: Core IT for geeks and pros
DBA Tools
ilonng
yangtingkun
NinGoo@Net
Oracle & Unix
Inside the Oracle Optimizer - Removing the black magic
Ricky's Test Blog
DBA@Taobao
存储部落
Think in 88

