Hi Guys,

I am a newbie to installing postgresql on linux as I'm a SQL DBA with postgresql on windows skills but now tasked with installing postgresql on RHCS and it seems there's no a-z way to do this from what I have researched thus far. All i'm looking for is a fundamental basic how to that i can follow and then i can subsequently learn how to twick to suite requirements. HELP

I put "how to install postgresql on red hat linux 6" into Google, and got this as my first hit:
https://wiki.postgresql.org/wiki/YUM_Installation

...an "a-z way" to install it. You also need to be aware that unless you are PAYING for RHEL, installing things through yum may be difficult, if possible at all. You should just be able to type in "yum list postgres*" to get a list of all postgres packages, then type in "yum install <package name>".

Thanks for the TBONE...I've had a look at the link and will try to digest the information on how to...
On another note is there a GUI way to carry out this install? Since this is going to be a RHCS install, isn't there some sort of a difference as to how the install is carried out when installing on a cluster?

1 members found this post helpful.

@TB0ne it seems I have a lot of reading up to do especially coming from a Windows background. What about the GUI aspect? Is this only achievable via package. For simplicity I was hoping as a beginner I would be able to install a GUI and use this to carry out the install and post install cluster configuration...I have installed postgresql before but only on Windows. postgresql 9.3 I think on windows 2008 R2.

The GUI aspect is easily done...there are already GUI based package tools available, but you REALLY need to learn the (simple) commands. While you CAN go into Windows control panel and click around for ten minutes to find out the IP address of an interface, isn't it easier to open a command-line and type in "ipconfig /all"?? This is much the same...typing in "yum install <package>" isn't too hard.
Installing it isn't difficult, see again the "yum install" command. Configuring it on a cluster is also not difficult. Again, RHEL 6 is OLD...are you PAYING FOR RHEL?? If so (as you should be), you can easily contact Red Hat support for assistance.

The command line is FAR more powerful than any GUI..Windows forces you to do things as THEY want to..Linux gives you options and tools.

Thanks TB0ne and I do agree with being able to use the command line but as a novice don't want to make any mistakes that might be detrimental to the project. Don't even know how to connect to linux to be able to administer it :-). I am sure the RHEL is paid for and might help to get support as you've said. I was just hoping I could get a newbie kind of how to connect to linux and sample install steps. I guess I have to spend time with using YUM commands. Is cluster install of postgreSQL different to non-cluster? If yes what are the things to look out for?
Lastly, streaming replication is also going to be used for DR and that is another one that i'm just trying to wrap my Microsoft head round :-)

As far as connecting goes, the std way is to install putty on your MS box and set that to use ssh to connect to the Linux box.
Obviously you need user/passwd credentials.
NB: for security, a Linux box is usually set to NOT allow remote logins as root (aka admin user), so you'll need a normal user acct and passwd, then after you have logged in you can
and supply the root user's passwd.
Presumably you have a Linux admin there - who setup the box?

On some systems, you would use
and supply your own passwd instead - again, ask the current admin.

This is a an old but gold intro to the cli - highly recommended http://rute.2038bug.com/index.html.gz

TB0ne I can't thank you enough as you've been greatly helpful and yes there are a few linux administrators here and I'm planning to use their expertise as much as possible.
On another note the whole environment is virtualized so not sure if I'll be connecting via windows or the virtual console which I presume eliminates the need for putty?

Another issue I might face is storage configuration. There's a requirement to split the location of volume groups into separate mount points e.g. Root_VG to /boot/swap, PGBIN_VG to /u01/BIN, PGDATA_1_VG to /u01/DB, PGWAL_1_VG to /u01/WAL and finally Archive_1_VG to /u01/AWAL. This I can do with my eyes closed in SQL Server but clueless as to how to do this in Linux. Not sure if this is specified during install or not and hence why I was hoping if I could have a GUI way of installing postgreSQL I might be able to configure this during the install? I just don't like doing something with little knowledge and giving up is never my way of life but I'm just a bit :-( right now...