First you need to set SELinux in permissive mode (if you have enabled it) and configure the firewall to let other systems to connect to port 80/443 (HTTP and HTTPs) and 389 (LDAP server port) using the commandline tool

# system-config-securitylevel-tui

Now we need to get some more RPM repos to work:

//First the CentOS testing repo
# cd /etc/yum.repos.d/
# wget http://dev.centos.org/centos/5/CentOS-Testing.repo
//Then the RPMForge repo
# rpm -Uhv http://apt.sw.be/redhat/el5/en/x86_64/rpmforge/RPMS//rpmforge-release-0.3.6-1.el5.rf.x86_64.rpm
//And last the GOsa RPM repo
# cat >> GOsa.repo << EOF
> [GOsa-Repository]
> name=GOsa Repository
> baseurl=ftp://oss.gonicus.de/pub/gosa/redhat
> enabled=0
> gpgcheck=0
> EOF

As we want the system as stable as possible without many packages from externarl repositories please verify that the files CentOS-Testing.repo and rpmforge.repo both contain the line ‘enabled=0‘. Now let’s check the new repositories for new packages to install and install them:

# yum –enablerepo=rpmforge,GOsa-Repository,c5-testing check-update
//Install LDAP
# yum install openldap openldap-servers
//Install PHP 5.2 from CentOS Testing
# yum –enablerepo=c5-testing install php-common php-imap php-snmp php-mysql php-mbstring php-pdo
//Install perl-Crypt-SmbHash from RPMForge
# yum –enablerepo=rpmforge install perl-Crypt-SmbHash
//Install GOsa RPMs
# yum –enablerepo=GOsa-Repository install gosa-help-en.noarch gosa-plugin-addressbook.noarch gosa-plugin-ldapmanager.noarch gosa-plugin-mail.noarch gosa-plugin-rolemanagement.noarch gosa-plugin-systems.noarch gosa-schema.noarch gosa-plugin-goto

Now we need to be sure that Apache HTTPD and LDAP servers start at boot:

# chkconfig httpd on
# chkconfig ldap on

At this point, before configuring GOsa, you need to add specific GOsa schemas to your ldap using as reference a part of the beginning of my own /etc/openldap/slapd.conf

include         /etc/openldap/schema/core.schema
include         /etc/openldap/schema/cosine.schema
include         /etc/openldap/schema/inetorgperson.schema
include         /etc/openldap/schema/gosa/rfc2307bis.schema
include         /etc/openldap/schema/gosa/gofax.schema
include         /etc/openldap/schema/gosa/gofon.schema
include         /etc/openldap/schema/gosa/samba.schema
include         /etc/openldap/schema/gosa/gosystem.schema
include         /etc/openldap/schema/gosa/goto.schema
include         /etc/openldap/schema/gosa/samba3.schema
include         /etc/openldap/schema/gosa/gosa-samba3.schema
include         /etc/openldap/schema/gosa/goserver.schema
include         /etc/openldap/schema/gosa/goto-mime.schema
include         /etc/openldap/schema/gosa/trust.schema

Restart your openldap and finally we can connect to GOsa web interface and configure it pointing your browser at

http://<hostname>/gosa/

Now you should have a fully functional GOsa install on your CentOS5: please let me know if you found error in this tutorial or even if you found it useful.

What is a MySQL cluster?

Let’s start explaining the architecture of a MySQL cluster with an image taken from dev.mysql.com:

As you may see the MySQL cluster is an aggregation of many components:

• one management server;
• many MySQL daemons that acts as “frontend”;
• many data nodes that store the real data.

This tutorial will guide you in the creation of a cluster with:

• one management node;
• two MySQL daemons;
• two data nodes.

Obviously you can expand this configuration simply adding the components you need. As stated at the beginning of this tutorial you can create all this setup on a single server (well you need 3 IPs on the server) or, and that would be very easy, you can split the MySQL cluster components on many servers.

On the linux machine, as root, you have to run the following commands:

//create user and groups
# groupadd oinstall
# groupadd dba
# useradd -g oinstall -G dba oracle
# passwd oracle

//prepare the filesystem
# mkdir /opt/oracle
# chown -R oracle:oinstall /opt/oracle

//install some needed packages
# yum install libXp gcc make setarch libaio glibc-devel

//change /etc/redhat-release as RHEL5 is not a supported platform
# cp /etc/redhat-release /etc/redhat-release.5
# echo redhat-4 > /etc/redhat-release

OK… as root we have now finished!

Now log in with the oracle user and download the Oracle Client 10g 64bit from Oracle website: once there get the file called 10201_client_linux_x86_64.cpio.gz. Open the cpio oracle client archive and run the client installation program:

# gunzip 10201_client_linux_x86_64.cpio.gz
# cpio -idmv < 10201_client_linux_x86_64.cpio
# cd client
# ./runInstaller

I pointed the oraInventory directory to /opt/oracle in the first screen; in the second one I choose to install the Oracle 10g Client in /opt/oracle/product/10.2.0/client_1/. Just before the installation finishes it prompts to you to execute two more commands as root: in my own installation I run

//execute the following commands as root
# /opt/oracle/orainstRoot.sh
# /opt/oracle/product/10.2.0/client_1/root.sh

but, as you may guess, your path may be different from mine.

Now the final step: we have to restore the /etc/redhat-release file we changed before:

# mv /etc/redhat-release.5 /etc/redhat-release

I hope this tutorial can help you… if you find errors please report them to me and I will correct as soon as possible.

On the linux machine, as root, you have to run the following commands:

//create user and groups
# groupadd oinstall
# groupadd dba
# useradd -g oinstall -G dba oracle
# passwd oracle

//prepare the filesystem
# mkdir /opt/oracle
# chown -R oracle:oinstall /opt/oracle

//install some needed packages
# yum install libXp gcc make setarch libaio glibc-devel

//change /etc/redhat-release as RHEL5 is not a supported platform
# cp /etc/redhat-release /etc/redhat-release.5
# echo redhat-4 > /etc/redhat-release

OK… as root we have now finished!

Now log in with the oracle user and download the Oracle Client 10g from Oracle website: once there get the file called 10201_client_linux32.zip. Unzip and run the client:

# unzip 10201_client_linux32.zip
# cd client
# ./runInstaller

I pointed the oraInventory directory to /opt/oracle in the first screen; in the second one I choose to install the Oracle 10g Client in /opt/oracle/product/10.2.0/client_1/. Just before the installation finishes it prompts to you to execute two more commands as root: in my own installation I run

//execute the following commands as root
# /opt/oracle/orainstRoot.sh
# /opt/oracle/product/10.2.0/client_1/root.sh

but, as you may guess, your path may be different from mine.

Now the final step: we have to restore the /etc/redhat-release file we changed before:

# mv /etc/redhat-release.5 /etc/redhat-release

I hope this tutorial can help you… if you find errors please report them to me and I will correct as soon as possible.

The  first thing is to get the HP-UX PA-RISC install server CDs from HP website:if you do not have the original CDs you can get them as trial from here.

Now its time to install the Disk Agent on the server you need to backup; you need to install a package that is not installed by default on the Ubuntu Linux 7.04: inetd. So, as “administrator”, execute the command:

> sudo apt-get install netkit-inetd

to get the needed package installed.

Insert the first CD of the HP-UX PA-RISC install server on the optical drive of the server and mount it:

> mount /dev/hda /mnt

Now it is install time:

> cd /mnt/LOCAL_INSTALL
> ./omnisetup.sh

Answer all the questions the installer asks to you and finish the installation. In my installation the Data Protector agent was correctly added to the inetd configuration:

> cat /etc/inetd.conf
omni stream tcp nowait root /opt/omni/lbin/inet inet -log /var/opt/omni/log/inet.log

OK… we are at the end…

> /etc/init.d/inetd restart

and now try if the Data Protector agent answers to port 5555:

> telnet localhost 5555

That’s all… if you encounter any problem leave a comment here… I’ll try to help you.

Leave a comment even if you have a smarter method… I am very glad to learn.

The  first thing is to get the HP-UX PA-RISC install server CDs from HP website:if you do not have the original CDs you can get them as trial from here.

Now its time to install the Disk Agent on the server you need to backup; you need to install two packages that are not installed by default on the Red Hat Enterprise Linux 5: ncompress and xinetd. So, as root, execute the command:

# yum install ncompress xinetd

to get the needed packages installed.

Insert the first CD of the HP-UX PA-RISC install server on the optical drive of the server and mount it:

# mount /dev/hda /mnt

Now it is install time:

# cd /mnt/LOCAL_INSTALL
# ./omnisetup.sh

Answer all the questions the installer asks to you and finish the installation. In my installation the Data Protector agent was not correctly added to the xinetd configuration… look if you find the omni file in /etc/xinetd.d… if you do not find it create it with the following content:

service omni
{
socket_type = stream
protocol = tcp
wait = no
user = root
server = /opt/omni/lbin/inet
server_args = inet -log /var/opt/omni/log/inet.log
disable = no
}

and verify it is owned by root and with 644 permissions.

OK… we are at the end…

# /etc/init.d/xinetd restart

As suggested by fluffy on comment #1 now you have to check if the firewall (iptables) is actually active: if the firewall is switched on you have to permit the traffic from and to port 5555 to flow regularly.

# chkconfig –list | grep iptables
iptables 0:off 1:off 2:off 3:off 4:off 5:off 6:off

If you any on you need to modify the file /etc/sysconfig/iptables adding:

-A RH-Firewall-1-INPUT -m state –state NEW -m tcp -p tcp –dport 5555 -j ACCEPT

Now everything should be ok… let’s try if the Data Protector agent answers to port 5555:

# telnet localhost 5555

That’s all… if you encounter any problem leave a comment here… I’ll try to help you.

Leave a comment even if you have a smarter method… I am very glad to learn.

Ovviamente il primo controllo è stato fatto sul DB (MySQL) e non ho trovato alcun messaggio d’errore… il problema è che in realtà il file /var/log/syslog era fermo da parecchie ore. Spinto da una irrazionale speranza ho provato un

/etc/init.d/mysql restart

per scoprire che era inutile… il DB non risaliva. Bene… ho i backup… ranzo tutto e rimetto su i backup: no… nemmeno questo era possibile… la directory dei backup non era più una directory e al posto dei famigliare drwxr-xr-x aveva degli spaventosi punti di domanda.

Riassumendo avevo il DB che non riusciva a risalire e i backup persi!

A questo punto l’unica cosa da fare era provare a verificare l’integrità del file system (ext3) e ho seguito i seguenti passi:

> init 1 #vai in single-user-mode
> mount #dammi la lista dei filesystem montati
/dev/sda1 on / type ext3 (rw,errors=remount-ro)
proc on /proc type proc (rw,noexec,nosuid,nodev)
/sys on /sys type sysfs (rw,noexec,nosuid,nodev)
varrun on /var/run type tmpfs (rw,noexec,nosuid,nodev,mode=0755)
varlock on /var/lock type tmpfs (rw,noexec,nosuid,nodev,mode=1777)
udev on /dev type tmpfs (rw,mode=0755)
devshm on /dev/shm type tmpfs (rw)
devpts on /dev/pts type devpts (rw,gid=5,mode=620)
> mount -o remount,ro / #rimonta in read-only il filesystem di /
> e2fsck -f -v /dev/sda1 #controlla il filesystem anche se sembra corretto (-f) e sii prolisso nei log (-v)
> mount -o remount,rw / #rimonta in read-write il filesystem di / una volta terminato il controllo
> init 2 #ritorna al runlevel precedente e alla piena operatività del sistema

Per fortuna dopo questa cura il DB è risalito correttamente e ho addirittura recuperato i miei backup!

Una volta scaricato il CD della Ubuntu 8.04 (qui la ALPHA-5) e masterizzato lo andiamo ad inserire nel nostro futuro server e provvediamo al restart dello stesso. Subito dopo la partenza saremo accolti da una schermata che chiede di scegliere la lingua d’installazione: visto che si tratta di un server io preferisco sempre installare in inglese.Scelto l’inglese viene subito proposta la schermata iniziale con il logo Ubuntu e dove la prima voce è “Install Ubuntu”; prima di scegliere questa voce dando INVIO sarà necessario prima premere F4 per poter scegliere l’opzione “Install a command-line system”. A questo punto potremo procedere con il premere INVIO e iniziamo così l’installazione.

• Dapper Drake: Ubuntu 6.04
• Edgy Eft: Ubuntu 6.10
• Feisty Fawn: Ubuntu 7.04
• Gutsy Gibbon: Ubuntu 7.10
• Hardy Heron: Ubuntu 8.04
• Intrepid Ibex: Ubuntu 8.10

Nonostante questa magnifica simmetria nei nomi a questa regola non rispondono però le prime releases:

• Warty Warthog: Ubuntu 4.10
• Hoary Hedgehog: Ubuntu 5.04
• Breezy Badger: Ubuntu 5.10

La mail originale di Mark Shuttleworth con il nuovo nome e gli intenti delle release 8.10 Intrepid Ibex li potete trovare sulla mailing list ubuntu-devel.