By and large, most Red Hat Linux systems will have Bash as the default shell. Bash is a darn great shell, but this article is about another equally great shell, called Z-Shell, that has most of the attributes of Bash, but in some cases goes the extra mile to give you the flexibility to customize your shell more than Bash allows.

This article is somewhat advanced, but if you’re very patient, with some effort, you will do just fine. Remember to make small changes, test them, and then make more small changes, test, and repeat.

First a few basics. I would encourage you to fire up a shell and test out this code while you read.

Basic configuration and setup

Switching your default shell to zsh

Before you start hacking away and changing config files, you need to do a few things first. Since Bash is the default shell, you will need to change things to Z-Shell, at least temporarily, while you follow this article. By far the easiest way to change your shell is to use the chsh command.

To quickly change to zsh, issue this command:

chsh -s /bin/zsh

If you take a peek at your /etc/passwd file:

less /etc/passwd 

You will see that the default shell for your user account has been changed. This is a permanent change, and when you log back in again you will stay with the zsh.

Z-Shell Configuration Files

The good and bad of it is that zsh can have as many as eight different configuration files. This is the order of execution that you should know about for the most common files:

1. /etc/zshenv - Always executed global configuration file.
2. ~/.zshenv - Local configuration file.
3. /etc/zprofile - Global Login shell configuration.
4. /etc/zshrc - Global interactive shell configuration.
5. ~/.zshrc - User interactive shell configuration.

The most efficient file to edit is probably ~/.zshenv as it will configure both interactive and non-interactive shells. For all of the examples I show below, please edit the ~/.zshenv file.

Backing up your configuration files or using version control

One thing I always caution is to backup your original file or use version control. One of the absolute worst situations to be in is with a damaged configuration file and no way to return to the original state. Make a backup of any file you edit, or better yet, use version control!

Z-Shell Cookbook

Since there are so many things you can do with Z-Shell and no way to talk about them all, I’m going to give you as much as possible through a cookbook format. I will go through the “recipes,” and you can “cook” along with your shell at home. Did you you remember to make a backup of your original files? Great, lets get started!

Autocompletion

Z-Shell has incredible autocompletion tools. Here are few barebones examples:

In order to use autocompletion, you will need to put the following in your ~/.zshenv file:

#Allows prompt profiles
autoload -U promptinit
promptinit

With zsh you can autocomplete things other than just file. If you type in:

$PA 

you get variable completion.

If you use something like gunzip in a directory with .gz files it will just figure out what files it should work with if you keep selecting tab. Try it out.

If you’re in the mood to get really geeky, check out completion functions in zsh. I don’t cover them in this article, but they are worth investigating.

Command history and history substitution

Mastering the use of command history will save you many keystrokes!

Remove mistakes:

root@cent]~# lesss /var/log/messages
zsh: command not found: lesss
[root@cent]~# ^s
less /var/log/messages

Remember commands:

[root@cent]~# ls /var/log/messages
/var/log/messages
[root@cent]~# !!
ls /var/log/messages
/var/log/messages
!$ Last Argument on the previous command line
cp !$ .
(copies /var/log/messages to current directory)

Rerun previous commands:

!l
root@cent]~# !l
ls /var/log/messages
/var/log/messages

This works because ls is in the history.

Finally, you can substitute by a command line number.

[root@cent]~# !32
ls /var/log/messages
/var/log/messages

Key bindings

Key bindings literally bind things to certain keys. Instead of using readline like Bash, zsh uses Zsh Line Editor or zle. In Bash you use bind and in zsh you use bindkey.

In zsh to find out what keys are bound:

bindkey -L

To see the whole list of editor commands that can be bound, and it is huge:

zle -la

I would recommend exploring keybindings by piping the list of your currently bound keys to less:

bindkey -L | less

Job Control

An example of using job control in Z-Shell:

tail -f /var/log/messages &
jobs
kill %1
run:

If you have a long-running process, you might want to let it work in the background while you continue other things. For example, if you download a tool I wrote called Liten, you may want to run it as a background job, as it will search a directory tree and do md5 checksums on files to find duplicates. This may take quite a while to completely get through a large file system, so it’s a perfect choice to run in the background.
Go ahead and run liten on your home directory as a background process:

./liten.py $HOME

Then click Ctrl+Z to send the job to the background.
The output will be:

zsh: suspended  ./liten.py $HOME

Then run:

jobs

And the output will be:

[1]  + suspended  ./liten.py $HOME
root@cent]~# bg %1
[1]  + continued  ./liten.py $HOME

If you have only one suspended job, you can use the shortcut:

%top
%vi 

You can also set an option in zsh to resume jobs by name:

setopt auto_resume

This lets you resume a job just by typing in the name of the job. It’s quite handy!

Remember that if you put a job in the background, you might want to pipe stdout somewhere while it runs.

When you exit the shell and you have a running job, zsh will kill the jobs by giving them a SIGHUP signal. Here are ways to tell zsh to prevent killing the jobs on exit:

• Use nohup in front of jobs you want to run on exit. Make sure it used in this manner:

  nohup top &

• You can also let your job run wild and give it the disown command or followup the the job with
  &!.

Priority of background jobs

In zsh, background jobs are automatically given a lower priority unless you set an option.

Process substitution

One neat feature zsh has is the special =(…) substitution.For example, if you want to find out the full path to a program you can do run:

ls =top

The output will be:

/usr/bin/top

I personally think this shorthand is quite nice, and I commonly find myself needing to find the full path to some command.

Another example would be when you want to use a shell command to format stdout and then
edit it in a text editor like vi:

vim =(tail < /var/log/messages)

Customizing Prompt

Like just about everything else in Z-Shell, the prompt is extremely customizable. If you would like to check out a few of the themes, you can set your .zshenv to allow prompt switching:

#Allows prompt profiles
autoload -U promptinit
promptinit

You can now view available prompt themes by typing:

prompt -l

You can also select and then set with the following:

prompt oliver
prompt -s

You will get back something like this:

prompt -s oliver
Set and save not yet implemented.

Please ensure your ~/.zshrc contains something similar to the following:

  autoload -U promptinit
  promptinit
  prompt oliver

If you want to get really fancy, you can also edit the following in your .zshenv file:

#Customize Prompt
PROMPT=$'[%n@%m][H:%B%!%b][J:%B%j%b]> '
RPS1='[D:%B%~%b]'

This is actually my current prompt, but feel free to tweak it around a bit; that’s how you learn. I won’t go any more into the details of customizing prompts here, but you should read through the Z-Shell manual.

File and directories

If you set:

setopt autocd

It will allow you to cd into a directory without typing cd first.

If you set:

cdable_vars

It will allow you to cd to a variable, such as:

cd HOME

Shortcuts for referring to directories

Get back to the directory you were just in:
~- same as $OLDPWD

Some fancy stuff:
~2 Second directory on the directory stack.
~-0 Final directory on the directory stack.

This works in conjunction with the pushd and the popd commands, which I won’t get into in this article, but you can read more here.

Using braces

Create ten test files in a few keystrokes:

touch {1..10}.txt

Summary

Z-Shell is a mind-blowing shell with the ability to tweak and change just about anything. What I covered in this brief survey article barely does Z-Shell justice. If you are interested in knowing more about Z-Shell or Bash, I would highly recommend reading, From Bash To Z-Shell by Oliver Kiddle, Jerry Peek, and Peter Stephenson. It is truly a masterful book! Of course, you can’t go wrong by reading the official documentation either.

Finally, I would love to hear back on some extra tricky things you have done with your Z-Shell.

©作者：Fenng 发布在 dbanotes.net

这个国庆跑了一大圈。9 月 25 日就开始休假了，加起来一共十几天的假期过得其实挺累。我的路线是：杭州－>上海－>成都－>黄龙－>九寨沟－>成都－>沈阳－>抚顺－>长春－>杭州。

先去的四川成都，然后走马观花了去了黄龙、九寨沟和都江堰。其实我是个比较懒的人，不过是答应了 Laura 今年要出来玩一次的，要讲究诚信嘛。

黄龙我个人觉得一般般。到了景点才知道为啥叫做黄龙。起这个名字的人一定是搞航拍的。

同行的还有一对印尼来的 Thomas 夫妇，导游的英语似乎不过关，我有的时候自告奋勇冒充一下翻译，可怜我的洋汀浜英语......这俩人行程安排的可倒挺满，基本把国内著名景点都逛遍了，挺佩服他们的体力。

九寨沟的水：

九寨沟还是名不虚传的，所谓"九寨归来不看水"倒是一点也不夸张。其实游客都是挺自私的，景观虽好，但不会因为你来了而变得更美，只会更糟。昨天看到世界遗产的中国悲歌这篇文章，觉得文中的担忧是对的。游客这么多，到哪里都和蝗虫似的。

没经得起司机的撺掇，也顺便去都江堰转了一圈：

大家都知道都江堰是李冰父子主导修建的。到底李冰的儿子叫啥? 这龙套跑的，好几千年了也没多少人知道。搜索了一下，似乎叫做”李二郎“，好像只是个小名啊。

这一圈很多时间坐车/飞机, 搞得我颈椎现在有反应了，今天不得不跑去做推拿。

更多图片请移步：一，二

--EOF--

相关文章|Related Articles

• 高尚生活
• 大雨杭州
• 在东北老家坐长途车的遭遇
• 三年前坝上木兰围场之行

评论数量(1)|Add Comments

本文网址：http://www.dbanotes.net/mylife/endless_journey.html

    Someone got their database lost today. A decision support system (DSS) database, 200GB size, running in noarchive log mode, and no backup available. One of the hard disk got corrupted, after got the data files back with disk utilities, they found there are a lot of corrupted or fractured blocks in system tablespace, about 100 damaged blocks found with DBVerify utility.

    I want to ask them, why no backup? Lack of storage? The disk space is cheap now, I can get 1TB storage within 500USD for PC hard disks, where they spent the storage money? Nobody know how to take effective backup? Hiring a DBA who knows how to perform backup is not expensive, where they spent the money to hire a qualified DBA? And they had no backup of the table structure, want to find these back from the corrupted system tablespace, it's not a easy job. Of cause, the cost saving is in wrong place.

    They asked me for possible recovery solutions. In this case, I told them that they can get some of the data recovered with Oracle DUL or AUL/MyDUL utility, if there are no corrupt blocks in other data files, we can get all the data recovered. But after ask me how I will charge them for the data recovery, they keep sliently, their department leaders are waiting for a free recovery. Is this correct cost saving method?

    A decision support system with 200GB data should be very important to the company, but they don't worry about it, seems the data is not important at all. Then why the company spend money in building it? Previous similar case is a OLAP database with 500GB data in it, finally they decide to spend several months of time to reload the data from business system because they do not want to pay for the recovery or they just want to pay several hundreds of dollars for it.

    We are very hard to get the reward for support service. A lot of costing saving occurred in wrong place.

Related Posts

Leave New Comment(Current: 0)

Link: http://www.dbatools.net/support/wrong_cost_saving.html

    I provide Oracle data recovery service for you, it's not free, so you will be charged, we need get an agreement on the charge first. How much you will be charged? I am not selling a product or a license, you may feel that I am just selling the AUL license, but I need to be online and provide support for you, so it's a support service actually. So there is no fixed price for this service. It varies on the database size and where are you from.

    But you need to tell me the company name and the phone number, please send these information via your enterprise email box. Else I may have problem to trust you.

    You do not have to process payment first before recovery. Because time saving is very important for recovery, everyone prefers get the data back first. Usually international payment may take several days to finish, can you wait several days? The most important is that I trust you, trust your promise. However, in China, I have to get the payment first.

    For international payment, I will send a bill invoice document to you to process the payment. The following information are critical for successful payment, ask your bank employee perform double check when processing.

1, Bank Name
2, Account Name
3, Account Number
4, SWIFT code of the bank

    Usually you should ask your finance department employee to process the payment after we successfully recover some data. And you should send me a image of the receipt from the bank, I will check whether the information is correct or not. The is required, if something wrong, we can fix it earlier.

    One of my customers wrote wrong account number, and I waited one month to get the payment. Another customer write wrong account name, and I waited near one month also to get the payment. I hope this won't happen again.

Related Posts

Leave New Comment(Current: 0)

Link: http://www.dbatools.net/mydul/aul_payment.html

因為Yahoo! Blog說系統升級，所以一整晚上不到，令我閱讀不到某些喜歡的Blog。Anyway，令所有用戶一整晚用不到，一定是大升級。那麼我只好懷著一些憧憬去等一下。

誰不知，再看看光復了的Yahoo! Blog。真的不知他們搞甚麼，也不知可說甚麼來形容。只覺得他們像Downgrade了整個服務。

沒錯，你們是加入了一些Ajax的介面，例如我們可以直接在一個blog的首頁，留言回應最新一篇的文章之類。不過，這類Ajax的小玩意，真的不能補償他們Downgrade了的用戶體驗。

首先，幹嗎每篇文都加一個醜樣的icon，算是甚麼呢？其實我第一次見到這icon，感到倒胃，於是立即去了其他網站。沒有icon不好嗎？有關這問題，不多說，因為阿Hang已詳細談論了。

很多Yahoo! Blog的用戶都將最新的回應列在首頁。「升級」之後，回應的次序亂了。我還以為我的舊文章有這麼多人回應。另外，「最新回應」所列出來的，只是某某人回應了某篇文章。為何不顯示回應的內文呢？如果這樣，不如直截了當的刪除這功能吧。

另外，「我的BLOG友」一欄強行把用戶的相片放大，令每個用戶的相都像打了馬賽克一樣，又是一大敗筆。

又另外，首篇文章的網誌日期，竟是XX個小時前。XX個小時前或XX天前算是甚麼日期呢？

其實我數的這幾點，都只是其中的Downgrade部份。他們還做了好些令用戶激憤的事情。大家可以閱讀他們《功能正逐步恢復》的官方文章，看看用戶們的回應有多澎湃便明白(15小時就有1182個回應，絕大部份是破口大罵的)。

另外on.cc亦開了一個討論，讓大家評價一下Yahoo! Blog的新版面。

P.S. on.cc像絕大部份的人，都不能正確的寫出Yahoo!的名稱。是「Yahoo!」，不是「Yahoo」呢。

Jan’s Tech Blog, where technologies are explored and annotated.

    When you prepare to do recovery of partitioned table with LOB columns with AUL4, you must modify the LOB index's partition name first in "AULOBJ.TXT". I have thought that the LOB index have the same partition name with table, just as normal local index, and in AUL 4 I use the table's partition name to locate the LOB index's partition to get the relative data object id. But the names are not same, let's start with two demo table :

SQL> CREATE TABLE T_HASHLOB (COL1 NUMBER, COL2 CLOB)
  2  LOB(COL2) STORE AS (DISABLE STORAGE IN ROW)
  3  PARTITION BY HASH(COL1) PARTITIONS 2;

Table created.

    Insert few rows and the perform a checkpoint, and start AUL 4 to unload the dictionary table, and describe the table structure:

AUL> desc anysql.t_hashlob

Storage(OBJ#=0 OBJD=0 TS=0 FILE=0 BLOCK=0 CLUSTER=0)
No. SEQ INT Column Name         Type
--- --- --- ------------------- ----------------
  1   1   1 COL1                NUMBER
  2   2   2 COL2                CLOB  (SYS_IL0000010046C00002$$)

    We could use grep to get the partition name information:

Read Full Text of【How AUL support the partition table's CLOB/BLOB recovery?】

Related Posts

Leave New Comment(Current: 0)

Link: http://www.dbatools.net/mydul/aul_lobpartition.html

dbanotes posted a photo:

    All the blocks of a chunk must be continual with a single extent, because you cannot specify a chunk size larger than the next extent size. Based on this rule, I add different chunk size support to AUL 4. Following are two demo tables, they both contains one column and only one row, populated with the same data.

SQL> SELECT TABLE_NAME,COLUMN_NAME,CHUNK FROM USER_LOBS;

TABLE_NAME      COLUMN_NAME       CHUNK
--------------- ------------ ----------
T_CHUNK2        COL1              32768
T_CHUNK1        COL1              16384

    Then we will use the chunk option of unload command to specify the LOB chunk size (how many blocks?) of this table, default is one. There is one limit that all the LOB columns should have the same chunk size in a single table. Then we start to unload the two LOB value to OS files :

Read Full Text of【How AUL support different Oracle CLOB/BLOB chunk size?】

Related Posts

Leave New Comment(Current: 0)

Link: http://www.dbatools.net/mydul/aul_lobchunk.html