Solaris

So here's the scoop. I am working on a project to automate the building of spark based solaris servers. I've got a cursory amount of knowledge of how to use the os after some pretty heavy experience with Linux, and doing a similar project there.

I have a few questions though...
1) how is hardware raid set up? In x86 this is usually handled via bios. Is it the same way in spark servers? Is it handled during installation?
2) successful installation of many Linux distress provides you with a kickstart file that you can use as a base for similar installs. Does a successful install of solaris provide you with a profile and rule file set to base jumpstart installs from?
I appreciate the help.
-Matty

There are 6 ways to install Solaris software. Two are interactive (GUI and CUI) and four are non-interactive:

- JumpStart
- WAN boot
- Solaris Flash
- Live Upgrade

There are pro and cons to each.

I suggest you read up about each of these and come back with questions when you have a basic understanding of the differences.

SPARC based ones actually. What models ?
Assuming you are going to install a recent Solaris 10 release or newer, I would recommend using ZFS based mirroring instead of hardware one.
In addition to fpmurphy reply, if you happen to use Solaris 11 Express, the installation process is quite different.
Jumpstart -> Automated Installer.
Live Upgrade -> Snap Upgrade/Boot Environment

Can you tell that first post was late at night? lol
I did mean SPARC based

I'm actually familiar with the install methods thanks to some install admin stuff we have here. Basically we're moving towards using a 3rd party provisioning tool for all of our server builds (Opsware, aka HP Service Automation). Its Solaris offering is based off of Jumpstart. Originally we had just been using jumpstart, so I'm hoping the move is easy. Basically opsware creates a skeleton of the jumpstart profile file and then allows certain fields to be populated dynamically so you can change configurations as needed (allowing different packages to be installed, different install clusters, etc).
I think what I need is a listing of what the most important fields are (the ones that are absolutely required for an installation).

As for the models we're going to be deploying, they're T5220's.

I would strongly suggest you do not use HPSA. What you call "creating a skeleton... allowing certain fields to be populated dynamically" really is "clumsy expensive GUI running jumpstart/WANboot in the background".

As for "minimum required fields"... you need to be more specific as to what that means, but:
For the later Solaris 10 releases, sysidcfg must have defined entries for:
keyboard
system_locale
timezone
terminal
timeserver
name_service
network_interface
security_policy
nfs4_domain

The minimum installation cluster would be SUNWCreq

Thanks for the listing of fields, I appreciate it. Sadly the choice about using/not using HPSA is out of my hands, and was made already; now its my job to implement it