Scalix Forums

Hi,
I tried to install Scalix on Centos 5.

1.) The given Information at: viewtopic.php?t=7849&highlight=centos are not correct (./scalix-installer --override-release=rhel5).
So I changed manually the Label in /etc/rethad-release to rhel5.
Who is giving this advise if it's not true (at least for version 11.0.4 GA)???

2) Then I tried to run the Scalix Installer but got this error:

The DHCP advise is o.k., no problem with that.
Now I tried to install manually the necessary packages, but then I was not able to get the full-install menu, so I deinstalled all Scalix packages.
But then the same advise appeared again.
What are the reason for this and how I can go around / avoid this issue??
Any ideas, hints?

Many thanks in advance

jochenthomas wrote:Hi,
I tried to install Scalix on Centos 5.

1.) The given Information at: viewtopic.php?t=7849&highlight=centos are not correct (./scalix-installer --override-release=rhel5).
So I changed manually the Label in /etc/rethad-release to rhel5.
Who is giving this advise if it's not true (at least for version 11.0.4 GA)???

2) Then I tried to run the Scalix Installer but got this error:

Code: Select all

 Environment check passed.
 Filesystem check passed.
 Network check warning -- network is configured for DHCP.
 Dependency check failed -- The following component(s) were missing from install/upgrade selection:
                            Scalix Server and Scalix Apache/Tomcat Connector.
                            Please return to the previous screen to select the missing component(s).
 Dependency check failed -- Please resolve the following package conflicts and re-run the system checks:
                            scalix-tomcat-connector is needed by scalix-tomcat-5.5.16-263.noarch
 Running services check passed.


The DHCP advise is o.k., no problem with that.
Now I tried to install manually the necessary packages, but then I was not able to get the full-install menu, so I deinstalled all Scalix packages.
But then the same advise appeared again.
What are the reason for this and how I can go around / avoid this issue??
Any ideas, hints?

Many thanks in advance

It looks like you dont have the redhat-release file right. The format changed in Redhat 5. It got me the first time i tried.

every package in the tar file is "noarch", except for scalix-server and scalix-tomcat-connector which it complained about. Those are specific to rhel5, which it didn't pick up because it didn't read the redhat-release file as being a rhel5 system.

a rhel5 redhat-release file should have the following....

Red Hat Enterprise Linux Server release 5 (Tikanga)

different from earlier versions...

Red Hat Enterprise Linux ES release 4 (Nahant Update 4)

Well, the advise about --override-release is true for Scalix 11.1 - we obviously recommend using the latest available version.

Florian.

Thanks for your answers.
Yes I used exact the mentioned phrase
"Red Hat Enterprise Linux Server release 5 (Tikanga)"
in my redhat-release file which I found in a thread at this forum (I was already goggling the forum).

@Florian, I downloaded the current version today.
Does you comment mean it is any more used in any newer version (and was only present at the 11.01)?

The only idea I have: Maybe I downloaded not the Red Hat Enterprise Linux 5.0 - tar.gz (85MB), so I will try again - maybe it's my fault

If not I will ask again, Thanks for your posts.

So, now I was downloading the right file (100 % sure: Red Hat Enterprise Linux 5.0) again.
Executed and saw a new problem:

Sorry to ask again: Do you have any ideas?
Is there anything missing, anything else?
Thanks in advance (again).

Well, i think the answer is right in the screen output - the installer tries to execute the OS-level 'dig' command and does not find the binary.

Obviously we are not checking for a mandatory depedency before start - however, the command is on most machines that we have installed.

On RHEL4, it seems that the command is part of the bind-utils package; I don't have a RHEL5 system at hand to check right now, but I would assume it's the same package that you would need to install.



Once confirmed, can you please log a bug in Scalix Bugzilla (bugzilla.scalix.com) against the installer that this will need to check for the presence of this package/binary before starting?

Thx,
Florian.

florian wrote:Well, i think the answer is right in the screen output - the installer tries to execute the OS-level 'dig' command and does not find the binary.

Obviously we are not checking for a mandatory depedency before start - however, the command is on most machines that we have installed.

On RHEL4, it seems that the command is part of the bind-utils package; I don't have a RHEL5 system at hand to check right now, but I would assume it's the same package that you would need to install.

Code: Select all

[florian@cp2 ~]$ rpm -q --file /usr/bin/dig
bind-utils-9.2.4-24.EL4

Once confirmed, can you please log a bug in Scalix Bugzilla (bugzilla.scalix.com) against the installer that this will need to check for the presence of this package/binary before starting?

Thx,
Florian.

confirming that in centos5/rhel5 its also the bind-utils package.

# rpm -q --whatprovides /usr/bin/dig
bind-utils-9.3.3-8.el5

Install bind-utils and you should be good to go.

Hi,
thanks for your response, I am only able to do this probably next weekend, but will report if this was the issue. I tried to find the package but without result...

Many thanks at this time.

jt___

same problem,
running with: OpenSUSE 10.2 (X86-64)
scalix-11.1.0-GA-core-suse101-intel.tgz

scalix-installer ran with override-release=suse101

edit: installed connector package by hand and it ist running now..


but i get an error when starting apache2:

# /etc/init.d/apache2 restart
httpd2-prefork: Syntax error on line 116 of /etc/apache2/httpd.conf: Syntax error on line 28 of /etc/apache2/sysconfig.d/loadmodule.conf: Cannot load /usr/lib64/apache2-prefork/mod_proxy_ajp.so into server: /usr/lib64/apache2-prefork/mod_proxy_ajp.so: undefined symbol: proxy_module