Got the NEW page working on the CRAB server……  It was more than likely a side effect of the Apache server upgrade somewhere around the Fedora 18 or 19 upgrade.  Here’s how I did it:

I became root user, then opened a terminal in the requests directory.

# nano .htaccess

Then, I put the following into the file:

Options +Includes
AddType text/html .html .shtml .txt
AddOutputFilter INCLUDES .html .shtml .txt

Saved the file, then restarted Apache as follows:

# systemctl restart httpd

That took care of it, but I think I shall try putting that .htaccess file in CRAB’s document root, as well as the main server document root.  I think that might be a better place for that file.  😀

Well, tonight, while trying to track down a firewall (I think) problem, I noticed that my messages log file, as well as all the other logs in /var/log/ were not being written, and were at 0 bytes.  Not very useful to tail.  So, I discovered that Fedora was moving away from rsyslog.  I was not happy.  So, I became root and:

# service rsyslog start

That took care of starting the logging back up, and to make sure that it starts after a reboot, I used nano to edit /etc/rc.d/rc.local and this is what I added towards the top of the file:

## Enable logging (Due to Fedora 20 suckiness)

service rsyslog start

THAT most definitely expresses my sentiment.  😉  I’m sure that uncovering these little issues is far from over, and all I can imagine is that Fedora is trying to ‘dumb down’ the OS to make it more like Windows, or Mac OS, in order to get a wider user base.  Might be what the average ‘Point and Click’ user wants, but NOT what I want.  Good thing that I’ve been at this game long enough to figure out how they are trying to tie my hands.  Well, now that I have the source port numbers, I can get back on the road to figuring out my problem, tomorrow.  😉

Last night, I discovered that http://crabradio.net/blogold/ wouldn’t load the index.php, and would only load a white screen.  Good thing that I decided to fix it this morning, because I discovered that the apache upgrade left the server ignoring .htaccess files, which allow for password protecting content and directories on the web server.  The result was a ‘wide-open’ web server.  It would only be a matter of time before Google bots and Yahoo spiders would start sucking me dry trying to download the contents of my web system.  Very similar to what happened when a bad symlink allowed open access to the entire filesystem.  Google bots were having a field day with that, and I’m sure would love it again, if I didn’t fix it right away.

Both items were a fairly simple fix in the same section of the /etc/httpd/conf/httpd.conf file.  (First, had to become root user.)

# nano /etc/httpd/conf/httpd.conf

I went to the Directory Directive for /var/www/html and added two lines highlighted in green…..

<Directory “/var/www/html”>
#
DirectoryIndex index.php
#
AllowOverride All

I actually edited the AllowOverride line from ‘None’ to ‘All’.  That line controls what directives can be placed into .htaccess files.  Since apache decided to make it None, it was ignoring them.  Thanks, Apache!  Last thing to do was restart or reload the httpd service.  This is done by:

# service httpd restart

But, since service has been replaced by systemctl, there is no telling when fedora will decide to fully depreciate service.  So, you should probably start getting used to the new command.  It does so much more, anyway.

# systemctl reload httpd

To make sure that it started okay:

# systemctl status httpd

or # service httpd status

Well, everything is now running as it should.  😉

One of the last things I had to do after imaging the system OS drive with Clonezilla, was to start a fresh backup with dar.  So, I manually started my custom script around midnight, and let it run.  When I woke up, I checked my email, and there wasn’t a backup email.  Curious.  So, I checked if it was still running.  Nope, was finished, but it choked and halted at the point of sending the email.

So, I discovered that there was no Mail Transport Agent installed, so HOW could the command ‘mail’ send any mail?  😉  So, using yum, I installed postfix.  Then, went through /etc/postfix/main.cf and using nano  (as root user) added a few basic things like hostname and domain.  I don’t think that there was much anything else that I had to edit.  So, time to start postfix.

# service postfix restart

The line in my backup script still refused to send out mail.  So, discovered that for some reason, postfix didn’t like my use of -a (attachment).  Had to change the line around like this:

mail -s Auto_Backup_SYS -a /mnt/storage/backupsys.txt -c >/dev/null

mail -s Auto_Backup_SYS -c < /mnt/storage/backupsys.txt >/dev/null

No big deal, since the status file is a text file, but I really don’t know why the -a flag was a show stopper.  While I was at it, I fixed a couple more vhosts for my friend, Don Landes.  Mental note – I really need to clean up the vhosts conf directory, as there are a few in there that aren’t used any longer.  That also means removing the ‘includes’ from the /etc/httpd/conf/httpd.conf file, but that can wait for another day.  😉