passing password to mysql
Stephan Wehner
password is read from a file, and should be passed to mysel. However, I
don't want to have the password appear in the list of processes (don't
use the mysql -p option)
A comment at the end of
http://dev.mysql.com/doc/refman/5.0/en/password-security-user.html
shows how to accomplish that within bash scripts, which is quite nice.
It works like the string in this code:
exec <<-END
{ printf
'[client]\ndatabase=%s\nhost=%s\nsocket=%s\nuser=%s\npassword=%s'
'db-dev' 'localhost' '/var/run/mysqld/mysqld.sock' 'db-user' 'db-passwd'
|
3<&0 <&4 4<&- /usr/local/mysql/bin/mysql --defaults-file=/dev/fd/3
} 4<&0
END
As you can see some clever use of file-descriptors. The exec-call works
in Ruby, and the mysql prompt is opened, but running "ps" will reveal
the password (not that surprising).
Is there a way to do this in Ruby?
Stephan
--
Posted via http://www.ruby-forum.com/.
Phlip
Grab ActiveRecord (part of Rails, ob-vossly), and write its database.yml file.
Oh, and actually use it to rip all your tables. It's just too easy to use!
Stephan Wehner
Ok thanks; however I find the mysql prompt sometimes easier and faster.
Not always but often enough.
Phlip
> Phlip wrote:
>>> Is there a way to do this in Ruby?
>> Grab ActiveRecord (part of Rails, ob-vossly), and write its database.yml
>> file.
>> Oh, and actually use it to rip all your tables. It's just too easy to
>> use!
>
> Ok thanks; however I find the mysql prompt sometimes easier and faster.
> Not always but often enough.
If you mean raw SQL statements, just .execute() them. Otherwise, how does the
MySQL connection driver work? It has to solve your actual problem on the inside,
right?
Stephan Wehner
Yes, I am aware of using connection.execute(..), select_all(...), etc.
However, I would like to be able to use the mysql prompt.
So is this a case where a bash script can do more than a Ruby script?