This is intended to document the steps involved in moving from an RCS based environment to a git based environment. In previous posts, I showed the environment I’m currently working in and the progress to changing from using RCS to using git. I duplicated the process in installing code to the target servers so that I can now use git from the command line very much like I used rcs and my scripts. Now I want to move a bit farther into the devops side of things, learning about setting up an easier way for the team to manage the scripts and code, providing a method for external teams to access some of the code, and replacing my sync scripts with Jenkins.
I’ll need to duplicate my environment again, this time with even the same server names. Since I have an ESX host at home, I can set up several servers that effectively match my official environment.
* git server: lnmt1cuomgit1
* Jenkins server: lnmt1cuomjkns1
* Target Tool server: lnmt1cuomtool1
* Target Status server: lnmt1cuomstat1
* Target Inventory server: lnmt1cuominv1
* Worker server: This will be a local, virtual box system where I can pull and test scripts and the two websites.
Again, this is partly to help me manage my personal sites better, but also to help with the work effort to get us more focused on Dev type tasks.
The Source server will have gitlab and Jenkins installed since Jenkins uses port 8080, they both should be able to cohabitate. Update: And after testing, that’s a big nope. So two different servers.
At least for the php site (Inventory), I’ll need to do a bit of rewriting as the settings.php file also contains the passwords to the database. Either the settings file needs to be replaced by a table or the passwords will need to be pulled from an external file outside of the site directory structure. Main bits right now though are the scripts the team actually uses. The Inventory has report scripts, data uploads, and of course the front end to the database which the scripts don’t touch.
On to gitlab…