To do this you will need Webmin installed so if you dont already go install it now. Here are the instructions: Install Webmin on OS X

To set up a Postfix monitoring system:

1. Open Webmin

2. Click the “others” tab

3. Click the “System and Server Status” icon

4. In the drop down box select “Postfix Server” and click “Add monitor of type”

5. Now you will now see on screen a bunch of options for your postfix monitor. The first setting you should change is “Check on schedule?”. Select in the drop down menu “Yes and report when going down”.

The next thing do is type your email address in the “Also send email for this service to” field. That section didn’t work for me so we will also send the email alerts with a different method. If they both work for you you can disable one of them.

6. Now comes the really cool part. Towards the bottom of the screen you will see an option under “Commands to run” called “If monitor goes down, run command”. This section allows you to run custom commands if postfix crashes. What we will do here is run a command that restarts Postfix then one to email you and tell you that Postfix has crashed.

In the “If monitor goes down, run command” field type:

sudo postfix start; tail -20 /var/log/mail.log | mail -s”mail log from postfix crash” youremail@youremailhost.com; tail -20 /var/log/system.log | mail -s”Postfix has crashed” youremail@youremailhost.com

Now as soon as your monitor detects a crash this system will restart postfix with the sudo postfix start; command and email you to tell you. It will also send you the last 20 entry’s in your mail log. The mail log entry’s will help you determine what went wrong. Make sure you replace youremail@youremailhost.com with your email address.
If you want in the “If monitor comes up, run command” type: echo ” Postfix back up ” | mail -s youremail@youremailhost.com youremail@youremailhost.com

That will send you an email that tells you that postfix was successfully restarted.

7. Click Save

Thats it your done. Now Postfix will monitor its self and restart if it crashes. If you the monitor is working right, on the System and Server Status page you will see this:

If Postfix is down you will see an X instead of the check mark.

You can also add any other custom commands to the “If monitor goes down, run command” just make sure you separate them by a ; Another command that might be useful is:

tail -20 /var/log/system.log | mail -s”system log from apache crash” youremail@youremailhost.com;

That will mail you the last 20 entry’s from your system log just before the crash. That will help you figure out what caused the crash.

Before doing this my Postfix server did not stay up longer then one day. Now Postfix has been running for over a week. It did crash a few times but my monitor started it right back up. I know I should probably figure out why it is crashing in the first place but until I do at least it works!

I tried calling apple and their only suggestion was to format the drive. Since this was my server I dreaded this option. Luckily before I reported the brutality of a reformat I did a little creative googling. I found a few different forum posts that when used together saved my sanity..

The first command which found where the problem files were is:

In the terminal type (this command takes a long time to complete):

sudo /usr/bin/find -x / -size +40000000c -exec /bin/ls -lh “{}” \;

That command will find and list all the large files on your hard drive.

In my case the problem was a bunch of MYSQL Replication bin log files that were over 1 gig each.

To delete the old mysql_binary_log files I ran this command in mysql:

Login to mysql: mysql -u username -p

Then type

PURGE MASTER LOGS BEFORE DATE_SUB( NOW(), INTERVAL 5 DAY);

That command will remove any bin that are older then 5 days.

Running that one line command save me countless hours of work.

I has a similar problem with my relay logs on one of my slave servers.
running the command:
sudo /usr/bin/find -x / -size +40000000c -exec /bin/ls -lh “{}” \;

showed me that my problem was with my relay logs.

Im not sure that this is the right way to fix the relay log problem but it worked in my situation:

type in mysql:

change master to master_host='’”;

This changes the master host and deletes the old logs. This may not be how you want to do this because you might need restart replication.

[Wed Jul 4 23:17:05 2007] [error] PHP Warning: ob_start() [ref.outcontrol]: output handler ‘ob_gzhandler’ conflicts with ‘zlib output compression’ in /script.php on line 419

I fixed it by adding this to my .htaccess file:

php_flag zlib.output_compression Off

I think the reason it was happing was that one of my scripts was trying to turn on compression when it was already on..

Another option would be to open the script causing the problem and edit out the lines that call compression.

It should look like:

Apple Feel free to use the fixed image, all I will charge you is a free iPhone ;->

So what does this mean?
Nothing really but maybe just maybe it means the maps program was not finished and the final will have GPS… Well at least I can dream….

1 - Open a Terminal window (Applications -> Utilities - > Terminal)

Navigate to the folder of the script you are trying to install.

Type:

cd path/folder

2 - Now Type:

Makefile.pl

or (depending on the script)

./configure

Hit Enter. The script will run through a lot of code and some time stop to ask you questions. To answer questions with the default answer just hit enter. For some files you might need to add parameters to the configure command like:

./configure –with-libpng=/sw \ –with-gd

Read the installation notes for more information.

3 - Next Type:

make

Hit enter.

4 - Type:

make test

5 - And Lastly type:

sudo make install

Hit enter, and enter you password.

Im in the process of switching to a new server. Tonight I noticed in my error log:

/var/log/httpd/error_log

This error:

[Mon Apr 23 02:27:30 2007] [error] PHP Warning: include(/script.php) [function.include]: failed to open stream: no suitable wrapper could be found in /script.php on line 26

I search around a bit and found the problem. The default instalation turns Has includes set to off for sucrity reasons. If you want to enable includes open:

/usr/local/php5/lib/php.ini
or
/usr/local/php/lib/php.ini

On Snow Leopard it is in:
/etc/php.ini

and search for:

Whether to allow include/require to open URLs (like http:// or ftp://) as files.
allow_url_include = Off

Switch to:

allow_url_include = On

Save and restart apache.

You can also turn on fopen from the same file, its right above the includes lines.

My text editor was unable to open the php.ini file because it did not have safficiant privileges. I used the editor in Webmin, if you have not installed Webmin yet you can edit it using pico

/usr/local/mysql/bin/mysqldump -u username -ppassword –skip-lock-tables –add-drop-table –all-databases > remotebackup.sql

I was using the Command Shell in Webmin

Usually when I backup using this method it saves the file to my user home folder located at:

Users/Username/

I viewed that folder but the file was no where in sight.
I went back to the command shell and typed ls to see the folders contents.
The files it returned were most of the same files it should have plus the files I was looking for.. I noticed though that folder did not have some of the files in my real home folder.

I finally found a terminal command that shows the path of the folder you are viewing. the command is PWD

PWD = Print Working Directory

I typed that into the shell and to my surprise I got:

/private/var/root

Im not sure why it got sent there, but at least I found it….

Here is a list of some of Webmins features:

-Configure almost all Apache directives and features.
-Create and edit domains and DNS records.
-Perform manual or scheduled backups and restores of configuration files managed by Webmin modules.
-View reports on bandwidth usage by host, port, protocol and time on a Linux system

-Bootup and Shutdown **(this allows you to restart you server remotely (from anywhere in the world))**
-Change the password of any user on the system.
-Schedule the transfer of files from this server to multiple servers in a Webmin cluster.
-Execute shell commands and view their output.
-View, edit and change permissions on files and directories on your system with a Windows-like file manager.
-Backup and restore filesystems using the dump and restore family of commands.
-Setup databases, tables and permissions in your MySQL database server.
-Configure PHP settings for the whole system or invidivual Apache virtual servers.
-Configure the Postfix mail server.
-Manage databases, tables and users in your PostgreSQL database server.
-List, kill and renice running processes on your system.
-Setup the SSH server for remote secure logins.
-Easily Create, edit and delete Cron jobs.
-Configure the syslog server on your system and view its log files.
-Upload multiple files to the server, and download multiple URLs either immediately or in the background at a scheduled time.

Plus a ton more..

you can see the full list at: http://www.webmin.com/standard.html

The OS X install page at http://www.webmin.com/osx.html requires you to install allot of stuff and is a little hard to follow.

The method im about to show you is much easier, but I can not guarantee all feature will work. I use this method all time and have not run into any problems.

1. First you will need to install Apple Developer Tools. Go to https://connect.apple.com/ and click Join Now to become a member.
Once you have registered log in then click Downloads then Developer Tools.
Download and install the latest release of Xcode

2. Next Download and decompress the latest version of Webmin from:
http://sourceforge.net/project/showfiles.php?group_id=17457

3. Rename the folder Webmin (removing the version number).

4. Drag the Webmin folder to you applications folder.

5. Open a Terminal window (Applications -> Utilities - > Terminal)

6. Type in terminal:

cd /Applications/webmin

7. Type in terminal:

sudo ./setup.sh

– You will be asked a bunch of questions, most of them will be configured automatically.
Type yes (Y) when asked whether to start webmin automatically at start up.
Hitting enter with out typing Y or N will choose the default value for the question.
Make sure you set your username and password during this configuration.

8. Once the setup is complete goto your browser and goto

http://127.0.0.1:10000
or
http://localhost:10000
or
http://yoursite.com:10000

To have mysqlcheck check and repair all of your databases:

Open a Terminal window (Applications -> Utilities - > Terminal)

Type:

mysqlcheck -u username -p -r –all-databases

enter your password when prompted, hit enter.

If you want to use mysqlcheck on just one database type:

mysqlcheck -u username -p -r databasename

The other day suddenly MYSQL on my OS X web server stopped working.
I was getting this error:
ERROR 2002 (HY000) Can’t connect to local MySQL server through socket /tmp/mysql.sock

No matter what I did all I got was:

Can’t connect to local MySQL server through socket ‘/tmp/mysql.sock’

after a few hours of searching I finally found the answer. Something had overwriting my /tmp folder. The tmp folder in OS X should be a symbolic link to /private/tmp.

To find out if your tmp file is a symbolic link

Open a Terminal window (Applications -> Utilities - > Terminal)

Type:

ls -la / | grep tmp

You should get:

lrwxr-xr-x 1 root staff 12 Nov 24 23:46 tmp -> /private/tmp

In my case it was not, it was an actual folder. To correct this:

Open a Terminal window (Applications -> Utilities - > Terminal)

Type:

sudo mv /tmp /tmp.old

Then type:

sudo ln -s /private/tmp /tmp

Reboot and restart MYSQL.