Unix vs. Windows Crib Sheet
Purpose
This document is intended to assist those
who are more familiar with a Windows environment in transitioning to a Unix AIX
environment. It should be considered a living document and additions/amendments
are welcomed from anyone who has further information or better ways of
accomplishing a task.
Accessing the System
Unless you have direct console access to
the machine you will need to control it remotely. Whereas with windows you may
be used to using tools such as remote desktop or terminal services to access a
server, with a Unix server you will need to use telnet or SSH to logon to the
server remotely.
Using Telnet
At the command prompt or in the run window
type: telnet <computer address>
e.g. telnet 10.3.1.12
Then a login prompt should appear into
which you can enter your login name and password
When you have finished using telnet type exit to leave the program.
Using SSH
While Telnet can be appropriate to use when
you are operating behind a firewall, SSH should always be used on an open
network and can be used anywhere. SSH encrypts all data transferred to the
server, including your password, unlike the older telnet protocol.
The easiest way to use SSH is to download
the SSH version of PuTTY from http://www.chiark.greenend.org.uk/%7Esgtatham/putty/download.html
PuTTY is a windows client to which allows
you to access a remote Unix machine via SSH (or Telnet or Rlogin). It also has various
options which can help you use the command line interface more effectively. The
complete documentation is available at:
General
-
AIX is case-sensitive. Best practice is to keep all filenames in
lower-case
-
AIX commands take switches and
the switches must be listed directly after the command
-
Use Escape K (Esc + K) to repeat the previous command
-
Use history –n to list the previous n
commands. Each command will have a
unique id listed. You can then repeat
one of these commands by entering r <id>
where <id> is the unique id
-
Previously entered commands can
be retrieved by pressing the Up arrow. For example, to retrieve the last but
one command entered press the Up arrow twice
-
Use Date to display or set the date or time
-
Whoami – The command that
displays who you are logged in as.
Cron Jobs
Cron jobs are time based commands often
used to trigger shell scripts to perform actions at a particular time, similar
to scheduled tasks in windows.
crontab
– l >> filename - This will export the cron jobs to a file, if the filename is not
specifiec (crontab –l) the list of files will be displayed on the
standard output e.g. the screen.
crontab
<filename> - This will import the cron jobs from a file. The crontab can also
be edited on screen but it is simpler to modify the cron jobs in file format.
Memory and Disk Usage
-
To monitor memory usage enter ps
aux | grep -v "PID %CPU %MEM"| grep -v kproc |sort +5 -rn | head -n
15
- du –am /Siebel/siebsrvr |
sort –rn > filesizes.txt – Will output the file
sizes sorted by largest first for all files/folder under the /Siebel/siebsrvr
tree to a filename filesizes.txt. This
is useful to work out which files are consuming the most disk space.
Help
man <command> – Will describe what the command does, its switches
and examples of how to use it
Navigating
ls – Will display a list of
all files and directories in your current location. Use switch –l to see
results in a list, the list will display date modified and read\write access
permissions as well as the file name.
cd – Will change your current directory: cd /Siebel/siebfile will take you to the Siebel
directory siebfile subdirectory (N.B. / indicates a directory in Unix rather
than \)
. - . indicates your current location .. indicates
the parent directory therefore cd .. will take you up one level in a directory
tree to the parent directory.
* - Is a
wildcard indicating one or more characters. This can be used with cd to save
the amount of typing where there is only one directory of a similar name. e.g.
cd /
pwd – Will print out the current working directory / show your current
location
File Management
cp – Will copy a file from
one location to another. You can use the –r option to recursively copy files,
i.e. copy whole directory structures. This can be used along with wildcards to
copy all files of a certain name:
cp df1.txt
./exampleA (Will copy the file named df1.txt in the current location to the
sub-directory named exampleA)
cp –r
./exampleA ./exampleB (Will copy the
directory named exampleA including all files and sub-directories to a directory
named exampleB)
cp
./exampleA/df* ./exampleB (Will copy all files in the directory exampleA whose
names begin with df to the directory exampleB)
scp
- Allows you to secure copy files between servers
on the same network. Generally, if you can ping the IP address of the
destination machine from the source machine you should be able to secure copy
between them.
scp filename
username@ip_address:/destination_path
eg. scp
in_fr_par.txt
sadmin@10.128.7.143:/siebel/siebfile/migration/mig_files/mig_fr_files
You will then be
prompted for the user’s password.
A status bar
will indicate progress of the copy.
mv – Will move a file from
one location to another. As with copy but the source file will not remain. This
can also be used for renaming a file.
mkdir – Will make a new directory with the provided name: mkdir
exampleA will create a new directory in the current location with the name
exampleA
chmod – Sets permissions on the file/folder. E.g. chmod 777 give R/W/E access to all
cat – Concatenates multiple
files into a single file or other output:
cat df1.txt
df2.txt > dfall.txt (Combines files df1.txt and df2.txt into a new file
dfall.txt)
rm - Removes /Deletes files
rm
df1.txt (Deletes the file df1.txt)
rm df*.txt (Deletes text files beginning with
df)
rm –r ./exampleA
(Deletes the directory exampleA and all it’s contents – USE WITH CAUTION)
rmdir – Removes empty
directories
Finding Data
To search files for a particular string you
can use the “grep” command. This has the following format:
grep
search_string filename
e.g grep error *.log
will return any line containing the word “error”
in a file with a name ending in .log. If the string being searched for contains
spaces then you need to put double quote marks ound the string
the grep command only searches in the
current directory. In order to search subdirectories you need to combine a find
command with grep e.g.
find
. -name *.log | xargs grep error
the grep command normally returns just the
line from the file where the data is found. However, it is possible to extract
more data than this using e “-p” switch (p = paragraph). This switch determines
the character(s) that mark the start and end of any line or paragraph. An
example of how this might be useful is as follows:
o If you have an XML file with a number of different entities but you
want to find some data that exists within a specific tag (such as
<activity>… </activity> ) and then write the complete tag and its
contents out to the file, you could use the following command:
grep -s -p"activity>" search_string filename >> out_filename
grep -s -p"activity>" search_string filename >> out_filename
Saving Results to a File
If you run a command then the results are
normally displayed on the screen; however the results can be saved to a file
using either the “>” command or the “>>” command.
“>” creates a new file (deleting any
existing file with the same name whereas “>>” appends to the file (or
creates it if it does not exist) e.g.
grep
search_string filename > out_filename
grep search_string filename >> out_filename
Compressing and Transferring Files
It is a common requirement to have to
download files from the UNIX server to your local pc where they can be more
easily examined. Such transfers can be done via the ftp command line program or
by using a specialist program such as SmartFTP which can be downloaded from http://www.smartftp.com/ which provides a
GUI for the transfer of files.
The time taken for such transfers can be
vastly reduced by compressing the file on the Unix server before transfer.
tar – The tar command can be used to create a single file which
contains a collection of files
tar cvvf
dffile.tar df1.txt (Would create a tar file containing the df1.txt file)
tar cvvf log.tar
*.log (Would create a single file named log.tar containing all files in the
current directory with the suffix .log)
gzip – Compresses files (similar to winzip)
gzip
-c dffile.tar > dffile.gz (would compress the tar file created previously)
gunzip – Uncompresses files (similar to winzip)
gunzip -c
dffile.gz > dffile.tar (would uncompress the previously compressed file)
More information on tar, gzip and gunzip
can be found at :
Siebel on Unix
File Structure:
In the typical Windows siebsrvr
installation, the log, archive, objects directories etc. are situated directly
within the siebsrvr directory. In Unix such directories should be accessed from
siebsrvr\enterprises\<enterprise_name>\<server_name>
This can be confusing as these directories
can also be found at the siebsrvr level.
Setting Environment Variables
Before using the command line srvrmgr
program or starting and stopping Siebel servers, you must ensure that the
environment variables are correctly set.
In the siebsrvr directory type .
./siebenv.sh
Starting and Stopping Siebel Processes
When you are starting and stopping
processes you can type:
ps –ef | grep sadmin
This will list all processes running under
sadmin, so that you can see whether or not the services are running. ps –ef | grep root can be used to check
for processes running under root which may in some environments include the web
processes.
Siebel
Server
In the siebsrvr/bin directory type:
To Stop: stop_server
ALL
To Start: start_server
ALL
Siebel
Gateway
In the gtwysrvr/bin directory type:
To Stop: stop_ns
To Start: start_ns
To View: list_ns
Web
Server
In the sweapp/bin directory type:
To Stop: .
./stopaps
To Start: .
./startapa
References
We also have a copy of the very useful UNIX
in a Nutshell on the bookshelf in the small meeting room in the London
