Discussion about CustomBuild

Why doesn't this tool support mod_security installation? It's quite common nowadays with apache
 
issue's again

Hi I have issue's similar to this thread http://www.directadmin.com/forum/showthread.php?t=25762&page=2

But this time it's a fresh install centos5.2 64bits

running setup.sh stops within seconds:
adding gd ...
--20:19:49-- http://files.directadmin.com/services/es_5.0_64/gd-2.0.33-9.3.fc6.x86_64.rpm
Resolving files.directadmin.com... 72.35.85.222
Connecting to files.directadmin.com|72.35.85.222|:80... connected.
HTTP request sent, awaiting response...


anyone any suggestions?
** Already solved myself :)***
solution was here: http://files2.directadmin.com/
 
Last edited:
Error establishing a database connection

Hi I just ran the custombuild script for the first time to update mysql etc and after it finished I got these errors and now all sites that need mysql are offline.


configure: error: Cannot find MySQL header files under /usr/local/mysql.
Note that the MySQL client library is not bundled anymore!

*** There was an error while trying to configure php. Check the configure/ap2/configure.php5 file

Any help would be great. :eek:

David
 
Nope, will not build.


./build mysql
--16:15:32-- http://files.directadmin.com/services/all/mysql/debian/mysql-4.1.22-linux-i686.tar.gz
=> `mysql-4.1.22-linux-i686.tar.gz'
Resolving files.directadmin.com... 72.35.85.222
Connecting to files.directadmin.com[72.35.85.222]:80... connected.
HTTP request sent, awaiting response... 404 Not Found
16:15:32 ERROR 404: Not Found.

Stopping mysqld ...
Stopping mysqld: [ FAILED ]
tar: mysql-4.1.22-linux-i686.tar.gz: Cannot open: No such file or directory
tar: Error is not recoverable: exiting now
tar: Child returned status 2
tar: Error exit delayed from previous errors
./build: line 2071: cd: mysql: No such file or directory
chown: cannot access `/usr/local/mysql-4.1.22-linux-i686': No such file or directory
/etc/init.d/mysqld: line 9: cd: /usr/local/mysql: No such file or directory
Giving mysqld a few seconds to start up...
cp: target `/usr/lib/mysql/' is not a directory: No such file or directory
Restarting MySQL.
Stopping mysqld: [ FAILED ]
/etc/init.d/mysqld: line 9: cd: /usr/local/mysql: No such file or directory
It is recommended to do "./build php n" after MySQL update.



David
 
Thank you for that, it's ok now. Thanks.

Apparently it was a corrupted mysql install because the file was not on DA's servers?

David
 
Last edited:
I ran the custom build (setup.sh -> 2. custom build....) in Centos 5.2 64-bit and everything seemed to work ok and I got the "setup finished" message after which there was a weird message.

HTML: 
*************************************
*   All parts have been installed   *
*************************************

Type: /sbin/service httpd restart
cp: `/var/www/html/squirrelmail/data' and `/var/www/html/squirrelmail-1.4.15/dat                  a' are the same file
./directadmin: error while loading shared libraries: libstdc++.so.6: cannot open                   shared object file: No such file or directory
./directadmin: error while loading shared libraries: libstdc++.so.6: cannot open                   shared object file: No such file or directory
After this, when I try to run DA I get this:
HTML: 
-bash-3.2# /etc/init.d/directadmin start
Starting DirectAdmin: /usr/local/directadmin/directadmin: error while loading shared libraries: libstdc++.so.6: cannot open shared object file: No such file or directory

libstdc++.so.6 is in /usr/lib64/ and if I run ldconfig:

HTML: 
-bash-3.2# ldconfig -p |grep libstd
        libstdc++.so.6 (libc6,x86-64) => /usr/lib64/libstdc++.so.6
it is really known by the system. What on earth should I do?

After a symbolic link in /lib/ I get the following

HTML: 
error while loading shared libraries: libstdc++.so.6: wrong ELF class: ELFCLASS64

Is this simply a matter of DirectAdmin executable being a 32-bit one?

EDIT: After loading the compatibility libraries (WHY ON EARTH don't the installation manuals mention this - I believed that the binary was really for CentOS-64) I finally got the system running but cannot log in. I am using the admin/PASS combination found in the setup.txt file but no luck so far. This is true pain.

My answer to the poll is that the custom build is far from a release version.

Running third setup in my virtual server now.

Markku
 
Last edited:
I use custombuild 1.1. httpd can not start. I get the same error in previous post in page 1.

Code: 
Starting httpd: Syntax error on line 29 of /etc/httpd/conf/extra/httpd-vhosts.conf:
The address or port is invalid

The line 29 from httpd-vhosts.conf

Code: 
<
VirtualHost :80>
    ServerAdmin webmaster@localhost
    AliasMatch ^/~([^/]+)(/.*)* /home/$1/public_html$2
    DocumentRoot /var/www/html
    ServerName localhost
    ScriptAlias /cgi-bin/ /var/www/cgi-bin/
    CustomLog /var/log/httpd/homedir.log homedir
</VirtualHost>

/etc/hosts file

Code: 
127.0.0.1               localhost.localdomain localhost
x.x.x.15           abc.mydomain.org abc

/etc/sysconfig/network file

Code: 
NETWORKING=yes
HOSTNAME=abc.mydomain.org
GATEWAY=x.x.120.1

I have tried to re build but it didn't help.
How to solve this!
 
mkniskanen said:
I EDIT: After loading the compatibility libraries (WHY ON EARTH don't the installation manuals mention this - I believed that the binary was really for CentOS-64) I finally got the system running but cannot log in. I am using the admin/PASS combination found in the setup.txt file but no luck so far. This is true pain.

My answer to the poll is that the custom build is far from a release version.

Running third setup in my virtual server now.
Click to expand...

Whew, it took me several hours to make DA work. The custom build process in a 64-bit system should definitely test the availability of the 32-bit-compatibility libraries BEFORE starting the process. Because this is neglected the process dies abruptly just before crossing the final line and the finalization is undone. The rebuild process taking such a very long time this is unforgivable.

Moreover, the test should tell the installer which libraries to install or it should be made very clear in the installation documentation. I do not think that this requirement is a very unrealistic one, given that the system actually tests the system stating: 'this must be a 64-bit system' or something alike.

It should actually say something like: "You are running a 64 bit system. Make sure you have the compatibility libraries (libstdc++...etc) installed"

Very, very, very un-professional from a piece of software we are paying for.

As for DA itself it is far below from what I expected but that is another story.
 
anhpt:
Code: 
<
VirtualHost :80>

Should be:
Code: 
<VirtualHost YOUR_IP:80>
 
Hello,

To: mkniskanen

It's most likely that the OS you've selected when you ordered was the 32-bit binary set. You do not need the 32-bit compatibility libraries for the DirectAdmin binaries if you select the 64-bit version of our software. Check your clients section to verify the OS you've selected. Let us know if it needs changing.

If you ever have issues with directadmin binaries themselves as you've descrived above, can can always send us an email as we've likely seen the problem before.

John
 
trouble when updating automake

Dauser2007 said:
Get error :

[root@host custombuild]# ./build update_versions
Updating automake
...
configure: error: Autoconf 2.60 or better is required.
*** The make has failed, do you want to try to make again? (y,n):
Click to expand...

Same problem here, at two new CentOS 5.2 boxes.
I just can update automake (./build automake) after changing the path:

"export PATH=${PATH}:/usr/local/bin"
 
Hello, I just made ./build update_versions and it vent well on two servers but on the third I now get a 500 Internal Server Error. :(

What can I do to get it to work again. :confused:
 
You must log in or register to reply here.
Back
Top