|
| |
|
In order to install Slack/390 Linux you must boot (IPL) a small version of it from a
tape or your z/VM virtual reader. The first file on that tape or in your virtual reader
will be the Linux/390 installation kernel. After that will be the kernel parameter
file, and then the initial ram disk that holds the root filesystem. Slack/390 Linux
comes with several installation kernel files from which you must choose one.
If you are going to be IPLing from tape, you must select the tape installation
kernel. If you are going to be IPLing from the z/VM virtual reader, you must select the
VM installation kernel. If you use the incorrect one, you will most likely experience
IPL errors. The kernel files are named so that it is clear which you should use.
IBM has published a Redbook on installing various Linux/390 distributions. While the
details are different for each distribution, the basic steps are the same. You can find
that book here:
Linux for zSeries and S/390: Distributions, SG24-6264-00. We strongly
recommend that you download it and read it. We'll re-cap the required steps here as well.
|
|
|
Creating The Boot (IPL) Medium
|
|
-
Download the appropriate installation kernel, the sample parmfile, and the initial
ramdisk (initrd) from the Slack/390 directory tree that corresponds to the version
you are installing. For example, the unstable/development version (-current),
would be at:
ftp://ftp.ibiblio.org/pub/Linux/distributions/slack390/slack390-current/kernels/
See the "Get Slack/390" page for all the
currently available versions.
-
Customize the kernel parameter file to match your installation's requirements. This
would include such things as your system's:
- DASD device numbers
- Host name
- IP address
- Network mask
- Network interface type (CLAW, OSA-2, etc.) and device numbers
- DNS servers
- Etc.
- Write the installation kernel, parameter file, and initrd out to tape, or the
z/VM virtual reader. If using tape, create an unlabeled tape, using a logical record
length and blocksize of 1024 bytes. If using the z/VM virtual reader, the files must
be 80-byte card images.
|
|
|
Booting (IPLing) the Starter System
|
|
|
IPL from the tape device, or the z/VM reader. Note that if the tape device control
unit is being shared by another system or systems, that you will very likely have to
attempt the IPL a number of times, due to the system's IPL code getting control unit busy
indications, etc. This will obviously not be an issue in a z/VM guest, since the virtual
reader won't have those sorts of errors.
|
|
|
Establishing the Network Connection
|
|
After you boot from tape or the virtual reader, you will be asked a number of questions
about your network connection, unless you provided all that information in the kernel
parameter file. If you did not, you will be prompted for your system's:
- fully qualified domain name (FQDN)
- network interface name (eth0, tr0, etc.)
- channel device layer parameters for your network interface
- IP address
- network mask (netmask)
- broadcast address
- network address
- default gateway IP address
- DNS server(s) IP address(es)
- DNS search domain(s)
If you are using a point-to-point connection, such as a CTC, ESCON channel, virtual
CTC, or IUCV, you will also be prompted for the IP address of the "peer" on the other
end of the connection.
After you provide the information, the system will attempt to activate the
interface, and ping itself, the default gateway, and the DNS server(s). |
|
|
Loading the DASD Device Drivers
|
|
If the network setup was successful, you will then be prompted for your DASD device
numbers. You can allow the system to try to autodetect your DASD volumes, but this is
not recommended.
You can specify the DASD device numbers as individual addresses, ranges of
addresses, or a combination of both. For example:
- 300,301,400,500
- 300-301,400,500
- 300-303,400-403,600
All this is a very good reason to specify all this information in the
kernel parameter file, since this is just a prelude to actually doing anything to get
software installed.
|
|
|
