Jump to contentJump to page navigation: previous page [access key p]/next page [access key n]
documentation.suse.com / SUSE Linux Enterprise Micro Documentation / Deployment Guide / Installation preparation / Installation on IBM Z and LinuxONE
Applies to SUSE Linux Enterprise Micro 5.4

4 Installation on IBM Z and LinuxONE

This chapter describes the procedure for preparing the installation of SUSE® Linux Enterprise Micro on IBM Z. It provides all information needed to prepare the installation on the LPAR and z/VM side.

4.1 System requirements

This section provides basic information about the system requirements, MicroCode level, and software for IBM Z.

4.1.1 Hardware

SUSE Linux Enterprise Micro runs on the following platforms:

  • IBM zEnterprise EC12 (zEC12) (2827)

  • IBM zEnterprise BC12 (zBC12) (2828)

  • IBM z Systems z13 (2964)

  • IBM z Systems z13s (2965)

  • IBM z Systems z14 (3906)

  • IBM z Systems z14 ZR1 (3907)

  • IBM z Systems z15 T01 (8561)

  • IBM z Systems z15 T02 (8562)

  • IBM z Systems z16 A01 (3931)

  • IBM LinuxONE Emperor (2964)

  • IBM LinuxONE Rockhopper (2965)

  • IBM LinuxONE Emperor II (3906)

  • IBM LinuxONE Rockhopper II (3907)

  • IBM LinuxONE III LT1 (8561)

  • IBM LinuxONE III LT2 (8562)

  • IBM LinuxONE Emperor 4 (3931)

4.1.1.1 Memory requirements

Different installation methods have different memory requirements during installation. At least 1 GB of memory is recommended for the text-mode installation under z/VM, LPAR, and KVM. Installation in the graphical mode requires at least 1.5 GB of memory.

Note
Note: Memory requirements with remote installation sources

A minimum of 512 MB of memory is required for installation from NFS, FTP, and SMB installation sources, or when VNC is used. Keep in mind that memory requirements also depend on the number of devices visible to the z/VM guest or the LPAR image. Installation with many accessible devices (even if unused for the installation) may require more memory.

4.1.1.2 Disk space requirements

The disk requirements depend largely on the workload of your containers. Minimal requirements for SLE Micro are 12 GB. The recommended value is 20 GB of hard disk space.

4.1.1.3 Network connection

A network connection is needed to communicate with your SUSE Linux Enterprise Micro system. This can be one or several of the following connections or network cards:

  • OSA Express Ethernet (including Fast and Gigabit Ethernet)

  • HiperSockets or Guest LAN

  • 10 GBE, VSWITCH

  • RoCE (RDMA over Converged Ethernet)

The following interfaces are still included, but no longer supported:

  • CTC (or virtual CTC)

  • ESCON

  • IP network interface for IUCV

For installations under KVM, make sure the following requirements are met to enable the VM Guest to access the network transparently:

  • The virtual network interface is connected to a host network interface.

  • The host network interface is connected to a network that the virtual server will join.

  • If the host is configured to have a redundant network connection by grouping two independent OSA network ports into a bonded network interface, the identifier for the bonded network interface is bond0. If more than one bonded interface exists, it is bond1, bond2, etc.

  • A non-redundant network connection setup requires the identifier of the single network interface. The identifier has the following format: enccw0.0.NNNN, where NNNN is the device number of the desired network interface.

4.1.2 MicroCode Level, APARs, and fixes

Documentation about restrictions and requirements for this release of SUSE Linux Enterprise Server be found on IBM developerWorks at https://developer.ibm.com/technologies/linux/. We recommend to use the highest service level available. Contact IBM support for minimum requirements.

For z/VM, the following versions are supported:

  • z/VM 6.4

  • z/VM 7.1

  • z/VM 7.2

  • z/VM 7.3

Since it might be necessary to activate the VM APARs before installing the new MicroCode levels, clarify the order of installation with IBM support.

4.1.3 Software

When installing SUSE Linux Enterprise Micro via non-Linux–based NFS or FTP, you might experience problems with NFS or FTP server software. The Windows* standard FTP server can cause errors, so we recommend performing installation via SMB on these machines.

To connect to the SUSE Linux Enterprise Micro installation system, one of the following methods is required (SSH or VNC are recommended):

SSH with terminal emulation (xterm compatible)

SSH is a standard Unix tool that is present on most Unix or Linux systems. For Windows, you can use the Putty SSH client.

VNC client

For Linux, the vncviewer VNC client is included in SUSE Linux Enterprise Micro as part of the tightvnc package. For Windows, TightVNC is also available. Download it from https://www.tightvnc.com/.

X server

Find a suitable X server implementation on any Linux or Unix workstation. There are many commercial X Window System environments for Windows and macOS*. Some can be downloaded as free trial versions.

Tip
Tip: More information

Before installing SUSE Linux Enterprise Micro on IBM Z, consult the README file located in the root directory of the first installation medium of SUSE Linux Enterprise Micro. The file complements this documentation.

4.2 Preparing for installation

This chapter explains how to make the data accessible for installation, install SUSE Linux Enterprise Micro using different methods, and prepare and use the IPL of the SUSE Linux Enterprise Micro installation system. The chapter also provides information about network configuration and network installation.

4.2.1 Making the installation data available

This section provides detailed information about making the SUSE Linux Enterprise Micro IBM Z installation data accessible for installation. Depending on your computer and system environment, choose between NFS or FTP installation. If you are running Microsoft Windows workstations in your environment, you can use the Windows network (including the SMB protocol) to install SUSE Linux Enterprise Micro on your IBM Z system.

Tip
Tip: IPL from DVD

It is possible to IPL from DVD and use the DVD as the installation medium. This is very convenient if you have restrictions setting up an installation server providing installation media over your network. The prerequisite is an FCP-attached SCSI DVD Drive.

Note
Note: No installation from hard disk

It is not possible to perform installation from a hard disk by putting the content of the DVD to a partition on a DASD.

4.2.1.1 Using a Linux workstation or SUSE Linux Enterprise Micro DVD

You can use a Linux workstation in your computer environment to provide the installation data to the IBM Z installation process by NFS or FTP.

Important
Important: Exporting mounted devices with NFS

Exporting the file system root (/) does not automatically export the mounted devices, such as DVD. Therefore, you need to explicitly name the mount point in /etc/exports:

/media/dvd  *(ro)

After changing this file, restart the NFS server with the command sudo systemctl restart nfsserver.

Setting up an FTP server on a Linux system involves the installation and configuration of server software like vsftpd. Downloading the installation data via anonymous login is not supported, therefore you need to configure the FTP server to support user authentication.

4.2.1.1.1 SUSE Linux Enterprise Micro on DVD

The first installation medium of the SUSE Linux Enterprise Micro for IBM Z contains a bootable Linux image for Intel-based workstations and an image for IBM Z.

For Intel-based workstations, boot from this medium. When prompted, choose the desired answer language and keyboard layout and select Start rescue system. You need at least 64 MB RAM for this. No disk space is needed, because the entire rescue system resides in the workstation's RAM. This approach requires setting up the networking of the workstation manually.

For IBM Z, IPL your LPAR/VM guest from this medium as described in Section 4.2.4.1.2, “IPL from FCP-attached SCSI DVD”. After entering your network parameters, the installation system treats the medium as the source of installation data. Because IBM Z cannot have an X11-capable terminal attached directly, choose between VNC or SSH installation. Refer to Section 13.3, “Monitoring installation via VNC” or Section 13.4, “Monitoring installation via SSH” for more information. SSH also provides a graphical installation by tunneling the X connection through SSH with ssh -X.

Important
Important: ssh -X connections between different architectures

By default, recent versions of the X.org and Xwayland servers do not accept connections from clients on different architectures. If you connect to the IBM Z machine from a AMD64/Intel 64 workstation with ssh -X, you will likely see the error message: Prohibited client endianess, see the Xserver man page.

To enable X connections between different architectures, create the file /etc/X11/xorg.conf.d/99-byte-swapping.conf with the following content:

Section "ServerFlags"
    Option "AllowByteSwappedClients" "on"
EndSection

Restart your X.org or Xwayland server to apply the configuration change:

> sudo systemctl restart display-manager.service

4.2.1.2 Using a Microsoft Windows workstation

You can use a Microsoft Windows workstation on your network to make the installation media available. The easiest way to do this is to use the SMB protocol. Make sure to activate SMB over TCP/IP as this enables the encapsulation of SMB packages into TCP/IP packages. Find details in the Windows online help or other Windows-related documentation that covers networking.

4.2.1.2.1 Using SMB

To make the installation media available with SMB, insert the USB flash drive with SLE-15-SP4-Online-ARCH-GM-media1.iso into the USB port of the Windows workstation. Then create a new share using the USB flash drive's letter and make it available for everyone in the network.

The installation path in YaST can be:

smb://DOMAIN;USER:PW@SERVERNAME/SHAREPATH

Where the placeholders mean:

DOMAIN

Optional workgroup or active directory domain.

USER, PW

Optional user name and password of a user who can access this server and its share.

SERVERNAME

The name of the server that hosts the share(s).

SHAREPATH

The path to the share(s).

4.2.1.2.2 With NFS

Refer to the documentation provided with the third party product that enables NFS server services for your Windows workstation. The USB flash drive containing the SLE-15-SP4-Online-ARCH-GM-media1.iso medium must be in the available NFS path.

4.2.1.2.3 Using FTP

Refer to the documentation provided with the third-party product that is enabling FTP server services on your Windows workstation. The USB flash drive containing the SLE-15-SP4-Online-ARCH-GM-media1.iso medium must be in the available FTP path.

The FTP server that is bundled with certain Microsoft Windows releases implements only a subset of the FTP commands, and it is not suitable for providing the installation data. In this case, use a third-party FTP server that offers the required functionality.

4.2.1.2.4 Using an FCP-attached SCSI DVD drive

After you IPLed from the SCSI DVD as described in Section 4.2.4.1.2, “IPL from FCP-attached SCSI DVD”, the installation system uses the DVD as the installation medium. In this case, you do not need the installation media on an FTP, NFS, or SMB server. However, you need the network configuration data for your SUSE Linux Enterprise Micro, because you must set up the network during the installation to perform a graphical installation via VNC or by X.

4.2.1.3 Using a Cobbler server for zPXE

IPLing from the network requires a Cobbler server to provide the kernel, initrd, and the installation data. Preparing the Cobbler server requires the following steps:

4.2.1.3.1 Importing the installation data

Importing the media requires the installation source to be available on the Cobbler server—either from USB flash drive or from a network source. Run the following command to import the data:

> sudo cobbler import --path=PATH1 --name=IDENTIFIER2 --arch=s390x

1

Mount point of the installation data.

2

A string identifying the imported product, for example sles15_s390x. This string is used as the name for the subdirectory where the installation data is copied to. On a Cobbler server running on SUSE Linux Enterprise this is /srv/www/cobbler/ks_mirror/IDENTIFIER. This path may be different if Cobbler runs on another operating system.

4.2.1.3.2 Adding a distribution

Adding a distribution allows Cobbler to provide the kernel and the initrd required to IPL via zPXE. Run the following command on the Cobbler server to add SUSE Linux Enterprise Micro for IBM Z:

> sudo cobbler distro add --arch=s390 --breed=suse --name="IDENTIFIER"1 \
  --os-version=slemicro5.22 \
  --initrd=/srv/www/cobbler/ks_mirror/IDENTIFIER/boot/s390x/initrd3 \
  --kernel=/srv/www/cobbler/ks_mirror/IDENTIFIER/boot/s390x/linux4 \
  --kopts="install=http://cobbler.example.com/cobbler/ks_mirror/IDENTIFIER"5

1

Unique identifier for the distribution, for example SLE Micro 5.4 IBM Z.

2

Operating system identifier. Use sles15.

3

Path to the initrd. The first part of the path (/srv/www/cobbler/ks_mirror/IDENTIFIER/) depends on the location of the imported data and the subdirectory name you chose when importing the installation data.

4

Path to the kernel. The first part of the path (/srv/www/cobbler/ks_mirror/IDENTIFIER/) depends on the location of the imported data and the subdirectory name you chose when importing the installation data.

5

URL to the installation directory on the Cobbler server.

4.2.1.3.3 Adjusting the profile

Adding a distribution (see Section 4.2.1.3.2, “Adding a distribution”) automatically generates a profile with the corresponding IDENTIFIER. Use the following command to make a few required adjustments:

> sudo cobbler distro edit \
--name=IDENTIFIER1 --os-version=sles102 --ksmeta=""3
--kopts="install=http://cobbler.example.com/cobbler/ks_mirror/IDENTIFIER"4

1

Identifier for the profile. Use the string specified when added the distribution.

2

Operating system version. Distribution to which the profile should apply. Use the string specified with --name=IDENTIFIER in the importing step.

3

Option required for templating Kickstart files. Since it is not used for SUSE, leave it empty.

4

Space-separated list of kernel parameters. It must include at least the install parameter.

4.2.1.3.4 Adding systems

The last step is to add systems to the Cobbler server. This step must be performed for every IBM Z guest that should boot via zPXE. Guests are identified by their z/VM user ID (in the following example, the ID linux01). Note that the ID must be lowercase. To add a system, run the following command:

> sudo cobbler system add --name=linux01 --hostname=linux01.example.com \
--profile=IDENTIFIER --interface=qdio \
--ip-address=192.168.2.103 --subnet=192.168.2.255 --netmask=255.255.255.0 \
--name-servers=192.168.1.116 --name-servers-search=example.com \
--gateway=192.168.2.1 --kopts="KERNEL_OPTIONS"

The --kopts option allows you to specify the kernel and installation parameters that are usually specified in the parmfile. Specify the parameters using the following format: PARAMETER1=VALUE1 PARAMETER2=VALUE2. The installer prompts for missing parameters. For a fully-automated installation, you need to specify all parameters for networking, DASDs and provide an AutoYaST file. Below is an example for a guest equipped with an OSA interface using the same network parameters as above.

--kopts=" \
AutoYaST=http://192.168.0.5/autoinst.xml \
Hostname=linux01.example.com \
Domain=example.com \
HostIP=192.168.2.103 \
Gateway=192.168.2.1 \
Nameserver=192.168.1.116 \
Searchdns=example.com \
InstNetDev=osa; \
Netmask=255.255.255.0 \
Broadcast=192.168.2.255 \
OsaInterface=qdio \
Layer2=0 \
PortNo=0 \
ReadChannel=0.0.0700 \
WriteChannel=0.0.0701 \
DataChannel=0.0.0702 \
DASD=600"

4.2.1.4 Installing from a USB Flash Drive of the HMC

Installation of SUSE Linux Enterprise Micro on IBM Z servers usually requires a network installation source. If this requirement cannot be fulfilled, SUSE Linux Enterprise Server allows you to use the USB flash drive of the Hardware Management Console (HMC) as an installation source for installation on an LPAR.

To perform installation from the USB flash drive of the HMC, proceed as follows:

Important
Important: Configure network

Before starting the installation, specify a network configuration in linuxrc. You cannot do this via boot parameters, and it is very likely that you will need network access. In linuxrc, go to Start Installation, then choose Network Setup.

Important
Important: Linux system must boot first

Before granting access to the media in the USB flash drive of the HMC, wait until the Linux system is booted. IPLing can disrupt the connection between the HMC and the LPAR. If the first attempt to use the described method fails, you can grant the access and retry the option HMC.

Note
Note: Installation repository

The USB flash drive is not kept as an installation repository, as the installation is a one-time procedure. If you need an installation repository, register and use the online repository.

4.2.2 Installation types

This section describes SUSE Linux Enterprise Micro installation steps for each installation mode. When the preparation steps described in the previous chapters have been completed, follow the overview of the desired installation mode.

As described in Section 4.2.1, “Making the installation data available”, there are three different installation modes for Linux on IBM Z: LPAR, z/VM, and KVM guest installation.

Procedure 4.1: Overview of an LPAR installation
  1. Prepare the devices needed for installation. See Section 4.2.3.1, “Preparing the IPL of an LPAR installation”.

  2. IPL the installation system. See Section 4.2.4.1, “IPLing an LPAR installation”.

  3. Configure the network. See Section 4.2.5, “Network configuration”.

  4. Connect to the SUSE Linux Enterprise Micro installation system. See Section 4.2.6, “Connecting to the SUSE Linux Enterprise Micro installation system”.

  5. Start the installation using YaST and IPL the installed system. See Chapter 12, Installation steps.

Procedure 4.2: Installation overview of z/VM installation
  1. Prepare the devices needed for installation. See Section 4.2.3.2.1, “Adding a Linux guest using dirMaint”.

  2. IPL the installation system. See Section 4.2.4.2, “IPLing a z/VM installation”.

  3. Configure the network. See Section 4.2.5, “Network configuration”.

  4. Connect to the SUSE Linux Enterprise Micro installation system. See Section 4.2.6, “Connecting to the SUSE Linux Enterprise Micro installation system”.

  5. Start the installation using YaST and IPL the installed system. See Chapter 12, Installation steps.

Procedure 4.3: Overview of a KVM guest installation
  1. Create a virtual disk image and write a domain XML file. See Section 4.2.3.3, “Preparing the IPL of a KVM guest installation”.

  2. Prepare the installation target and IPL the VM Guest. See Section 4.2.4.3, “IPLing a KVM guest installation”.

  3. Section 4.2.5.3, “Set up the network and select the installation source”.

  4. Connect to the SUSE Linux Enterprise Micro installation system. See Section 4.2.6, “Connecting to the SUSE Linux Enterprise Micro installation system”.

  5. Start the installation using YaST and IPL the installed system. See Chapter 12, Installation steps.

4.2.3 Preparing the IPL of the SUSE Linux Enterprise Micro installation system

4.2.3.1 Preparing the IPL of an LPAR installation

Configure your IBM Z system to start in ESA/S390 or Linux-only mode with an appropriate activation profile and IOCDS. For further information, refer to the IBM documentation. Continue as described in Section 4.2.4.1, “IPLing an LPAR installation”.

4.2.3.2 Preparing the IPL of a z/VM installation

4.2.3.2.1 Adding a Linux guest using dirMaint

The first step is to attach and format one or multiple DASDs in the system to be used by the Linux guest in z/VM. Next, create a new user in z/VM. The example shows the directory for a user LINUX1 with the password LINPWD, 1 GB of memory (extendable up to 2 GB), several minidisks (MDISK), two CPUs, and an OSA QDIO device.

Tip
Tip: Assigning memory to z/VM guests

When assigning memory to a z/VM guest, make sure that the memory size is adequate for the preferred installation type as described in Section 4.1.1.1, “Memory requirements”. To set the memory size to 1 GB, use the command CP DEFINE STORAGE 1G. After the installation has finished, reset the memory size to the desired value.

Example 4.1: Configuration of a z/VM directory
USER LINUX1 LINPWD 1024M 2048M G
*____________________________________________
* LINUX1
*____________________________________________
* This VM Linux guest has two CPUs defined.

CPU 01 CPUID 111111
CPU 02 CPUID 111222
IPL CMS PARM AUTOCR
IUCV ANY
IUCV ALLOW
MACH ESA 10
OPTION MAINTCCW RMCHINFO
SHARE RELATIVE 2000
CONSOLE 01C0 3270 A
SPOOL 000C 2540 READER *
SPOOL 000D 2540 PUNCH A
SPOOL 000E 3203 A
* OSA QDIO DEVICE DEFINITIONS
DEDICATE 9A0 9A0
DEDICATE 9A1 9A1
DEDICATE 9A2 9A2
*
LINK MAINT 0190 0190 RR
LINK MAINT 019E 019E RR
LINK MAINT 019D 019D RR
* MINIDISK DEFINITIONS
MDISK 201 3390 0001 0050 DASD40 MR ONE4ME TWO4ME THR4ME
MDISK 150 3390 0052 0200 DASD40 MR ONE4ME TWO4ME THR4ME
MDISK 151 3390 0253 2800 DASD40 MR ONE4ME TWO4ME THR4ME

This example uses minidisk 201 as the guest's home disk. Minidisk 150 with 200 cylinders is the Linux swap device. Disk 151 with 2800 cylinders holds the Linux installation.

As user MAINT, add the guest to the user directory with DIRM FOR LINUX1 ADD. Enter the name of the guest (LINUX1) and press F5. Set up the environment of the user with:

DIRM DIRECT
DIRM USER WITHPASS

The last command returns a reader file number. This number is needed for the next command:

RECEIVE <number> USER DIRECT A (REPL)

You can now log in on the guest as user LINUX1.

If you do not have the dirmaint option available, refer to the IBM documentation on how to set up this user.

Proceed with Section 4.2.4.2, “IPLing a z/VM installation”.

4.2.3.3 Preparing the IPL of a KVM guest installation

A KVM guest installation requires a domain XML file that specifies the virtual machine and at least one virtual disk image for the installation.

4.2.3.3.1 Create a virtual disk image

By default, libvirt searches for disk images in /var/lib/libvirt/images/ on the VM Host Server. Although images can also be stored anywhere on the file system, it is recommended to store all images in a single location for easier maintainability. To create an image, log in to the KVM host server and run the following command:

qemu-img create -f qcow2 /var/lib/libvirt/images/s12lin_qcow2.img 10G

This creates a qcow2 image with a size of 10 GB in /var/lib/libvirt/images/.

4.2.3.3.2 Write a domain XML file

A domain XML file is used to define the VM Guest. To create the domain XML file, open an empty file s15-1.xml with an editor and create a file like in the following example.

Example 4.2: Example domain XML file

The following example creates a VM Guest with a single CPU, 1 GB RAM, and the virtual disk image created in the previous section (Section 4.2.3.3.1, “Create a virtual disk image”). It assumes that the virtual server is attached to the host network interface bond0. Change the source devices element to match your network setup.

<domain type="kvm">
 <name>s15-1</name>
 <description>Guest-System SUSE SLES15</description>
 <memory>1048576</memory>
 <vcpu>1</vcpu>
 <os>
  <type arch="s390x" machine="s390-ccw-virtio">hvm</type>
  <!-- Boot kernel - remove 3 lines after successfull installation -->
  <kernel>/var/lib/libvirt/images/s15-kernel.boot</kernel>
  <initrd>/var/lib/libvirt/images/s15-initrd.boot</initrd>
  <cmdline>linuxrcstderr=/dev/console</cmdline>
 </os>
 <iothreads>1</iothreads>
 <on_poweroff>destroy</on_poweroff>
 <on_reboot>restart</on_reboot>
 <on_crash>preserve</on_crash>
 <devices>
  <emulator>/usr/bin/qemu-system-s390x</emulator>
  <disk type="file" device="disk">
   <driver name="qemu" type="qcow2" cache="none" iothread="1" io="native"/>
   <source file="/var/lib/libvirt/images/s15lin_qcow2.img"/>
   <target dev="vda" bus="virtio"/>
  </disk>
  <interface type="direct">
   <source dev="bond0" mode="bridge"/>
   <model type="virtio"/>
  </interface>
  <console type="pty">
   <target type="sclp"/>
  </console>
 </devices>
</domain>

4.2.4 IPLing the SUSE Linux Enterprise Micro installation system

4.2.4.1 IPLing an LPAR installation

There are different ways to IPL SUSE Linux Enterprise Micro into an LPAR. The preferred way is to use the Load from CD-ROM or server feature of the SE or HMC.

4.2.4.1.1 IPL from DVD-ROM

Mark the LPAR to install and select Load from CD-ROM or server. Leave the field for the file location blank or enter the path to the root directory of the first DVD-ROM and select Continue. Keep the default selection in the list of options that appears. Operating system messages should now show the kernel boot messages.

4.2.4.1.2 IPL from FCP-attached SCSI DVD

You can use the Load procedure by selecting SCSI as Load type to IPL from SCSI. Enter the WWPN (Worldwide port name) and LUN (Logical unit number) provided by your SCSI bridge or storage (16 digits—do not omit the trailing 0s). The boot program selector must be 2. Use your FCP adapter as Load address and perform an IPL.

4.2.4.2 IPLing a z/VM installation

This section describes IPLing the installation system to install SUSE Linux Enterprise Micro for IBM Z on a z/VM system.

4.2.4.2.1 IPL from the z/VM reader

You need a working TCP/IP connection and an FTP client program within your newly-defined z/VM guest to transfer the installation system via FTP. Setting up TCP/IP for z/VM is beyond the scope of this manual. Refer to the appropriate IBM documentation.

Log in as the z/VM Linux guest to IPL. Make the content of the directory /boot/s390x of the Unified Installer (medium 1) available via FTP within your network. From this directory, get the files linux, initrd, parmfile, and sles.exec. Transfer the files with a fixed block size of 80 characters. Specify it with the FTP command locsite fix 80. It is important to copy linux (the Linux kernel) and initrd (the installation image) as binary files, so use the binary transfer mode. parmfile and sles.exec need to be transferred in ASCII mode.

The following example shows the required steps. This particular scenario assumes that the required files are accessible from an FTP server at the IP address 192.168.0.3 and the login is lininst.

Example 4.3: Transferring the binaries via FTP
FTP 192.168.0.3
VM TCP/IP FTP Level 530
Connecting to 192.168.0.3, port 21
220 ftpserver FTP server (Version wu-2.4.2-academ[BETA-18](1)
Thu Feb 11 16:09:02 GMT 2010) ready.
USER
lininst
331 Password required for lininst
PASS
******
230 User lininst logged in.
Command:
binary
200 Type set to I
Command:
locsite fix 80
Command:
get /media/dvd1/boot/s390x/linux sles.linux
200 PORT Command successful
150 Opening BINARY mode data connection for /media/dvd1/boot/s390x/linux
(10664192 bytes)
226 Transfer complete.
10664192 bytes transferred in 13.91 seconds.
Transfer rate 766.70 Kbytes/sec.
Command:
get /media/dvd1/boot/s390x/initrd sles.initrd
200 PORT Command successful
150 Opening BINARY mode data connection for /media/dvd1/boot/s390x/initrd
(21403276 bytes)
226 Transfer complete.
21403276 bytes transferred in 27.916 seconds.
Transfer rate 766.70 Kbytes/sec.
Command:
ascii
200 Type set to A
Command:
get /media/dvd1/boot/s390x/parmfile sles.parmfile
150 Opening ASCII mode data connection for /media/dvd1/boot/s390x/parmfile
(5 bytes)
226 Transfer complete.
5 bytes transferred in 0.092 seconds.
Transfer rate 0.05 Kbytes/sec.
Command:
get /media/dvd1/boot/s390x/sles.exec sles.exec
150 Opening ASCII mode data connection for /media/dvd1/boot/s390x/sles.exec
(891 bytes)
226 Transfer complete.
891 bytes transferred in 0.097 seconds.
Transfer rate 0.89 Kbytes/sec.
Command:
quit

Use the REXX script sles.exec you downloaded to IPL the Linux installation system. This script loads the kernel, parmfile, and the initial RAM disk into the reader for IPL.

Example 4.4: sles.exec
/* REXX LOAD EXEC FOR SUSE LINUX S/390 VM GUESTS       */
/* LOADS SUSE LINUX S/390 FILES INTO READER            */
SAY ''
SAY 'LOADING SLES FILES INTO READER...'
'CP CLOSE RDR'
'PURGE RDR ALL'
'SPOOL PUNCH * RDR'
'PUNCH SLES LINUX A (NOH'
'PUNCH SLES PARMFILE A (NOH'
'PUNCH SLES INITRD A (NOH'
'IPL 00C'

Using the script, you can IPL the SUSE Linux Enterprise Micro installation system with the command sles. The Linux kernel then starts and outputs its boot messages.

To continue the installation, proceed to Section 4.2.5, “Network configuration”.

4.2.4.2.2 IPL from FCP-attached SCSI DVD

To IPL in z/VM, prepare the SCSI IPL process by using the SET LOADDEV parameter:

SET LOADDEV PORTNAME 200400E8 00D74E00 LUN 00020000 00000000 BOOT 2

After setting the LOADDEV parameter with the appropriate values, IPL your FCP adapter, for example:

IPL FC00

To continue the installation, proceed with Section 4.2.5, “Network configuration”.

4.2.4.2.3 IPL from a Cobbler server with zPXE

To IPL from a Cobbler server with zPXE, you need to transfer the zpxe.rexx script via FTP from the Cobbler server to your z/VM guest. To do this, the z/VM guest needs a working TCP/IP connection and an FTP client program.

Log in as the z/VM Linux guest to IPL and transfer the script with a fixed size of 80 characters in ASCII mode (see Example 4.3, “Transferring the binaries via FTP” for an example). The zpxe.rexx script is available on the Unified Installer DVD at /boot/s390x/zpxe.rexx or on a SLE Cobbler server at /usr/share/doc/packages/s390-tools/zpxe.rexx.

zpxe.rexx is supposed to replace the PROFILE EXEC of your guest. Make a backup copy of the existing PROFILE EXEC and rename ZPXE REXX to PROFILE EXEC. Alternatively, call ZPXE REXX from the existing PROFILE EXEC by adding the 'ZPXE REXX' line to it.

The last step is to create a configuration file ZPXE CONF that instructs ZPXE REXX which Cobbler server to contact and which disk to IPL. Run xedit zpxe conf a and create ZPXE CONF with the following content (replace the example data accordingly):

HOST cobbler.example.com
IPLDISK 600

This connects the Cobbler server next time you log in to the z/VM guest. If an installation is scheduled on the Cobbler server, it will be executed. To schedule the installation, run the following command on the Cobbler server:

> sudo cobbler system edit --name ID1 --netboot-enabled 12 --profile PROFILENAME3

1

z/VM user ID.

2

Enable IPLing from the network.

3

Name of an existing profile, see Section 4.2.1.3.3, “Adjusting the profile”.

4.2.4.3 IPLing a KVM guest installation

To start the guest installation, you first need to start the VM Guest defined in Section 4.2.3.3.1, “Create a virtual disk image”. Before you begin, ensure the kernel and initrd are available for IPL.

4.2.4.3.1 Preparing the installation source

Kernel and initrd of the installation system need to be copied to the VM Host Server to IPL the VM Guest into the installation system.

  1. Log in to the KVM host and make sure you can connect to the remote host or device serving the installation source.

  2. Copy the following two files from the installation source to /var/lib/libvirt/images/. If the data is served from a remote host, use ftp, sftp, or scp to transfer the files:

    /boot/s390x/initrd
    /boot/s390x/cd.ikr
  3. Rename the files on the KVM host:

    > sudo cd /var/lib/libvirt/images/
    > sudo mv initrd s15-initrd.boot
    > sudo mv cd.ikr s15-kernel.boot
4.2.4.3.2 IPL the VM Guest

To IPL the VM Guest, log in to the KVM host and run the following command:

> virsh  create s15-1.xml --console

The installation process starts when the VM Guest is up and running, and you should see the following message:

Domain s15-1 started
Connected to domain s15-1
Escape character is ^]
Initializing cgroup subsys cpuset
Initializing cgroup subsys cpu
Initializing
cgroup subsys cpuacct
.
.
Please make sure your installation medium is available.
Retry?
0) <-- Back <--
1) Yes
2) No

Answer 2) No and choose Installation on the next step. Proceed as described in Section 4.2.5.3, “Set up the network and select the installation source”.

4.2.5 Network configuration

Wait until the kernel has completed its start-up routines. If you perform the installation in basic mode or in an LPAR, open the Operating System Messages on the HMC or SE.

First, choose Start Installation in the linuxrc main menu. Then choose Start Installation or Update to start the installation process. Select Network as the installation medium, then select the type of network protocol to use for the installation. Section 4.2.1, “Making the installation data available” describes how to make the installation data available for the various types of network connections. Currently, FTP, HTTP, NFS, and SMB/CIFS (Windows file sharing) are supported.

From the list of available devices, choose an OSA or HiperSockets network device for receiving the installation data. Although the list may contain CTC, ESCON, or IUCV devices, they are no longer supported on SUSE Linux Enterprise Micro.

4.2.5.1 Configure a HiperSockets interface

Select a HiperSocket device from the list of network devices. Then enter values for the read, write, and data channels:

Example 4.5: Supported network connection types and driver parameters
Choose the network device.

 1) IBM parallel CTC Adapter (0.0.0600)
 2) IBM parallel CTC Adapter (0.0.0601)
 3) IBM parallel CTC Adapter (0.0.0602)
 4) IBM Hipersocket (0.0.0800)
 5) IBM Hipersocket (0.0.0801)
 6) IBM Hipersocket (0.0.0802)
 7) IBM OSA Express Network card (0.0.0700)
 8) IBM OSA Express Network card (0.0.0701)
 9) IBM OSA Express Network card (0.0.0702)
10) IBM OSA Express Network card (0.0.f400)
11) IBM OSA Express Network card (0.0.f401)
12) IBM OSA Express Network card (0.0.f402)
13) IBM IUCV

> 4

Device address for read channel. (Enter '+++' to abort).
[0.0.0800]> 0.0.0800

Device address for write channel. (Enter '+++' to abort).
[0.0.0801]> 0.0.0801

Device address for data channel. (Enter '+++' to abort).
[0.0.0802]> 0.0.0802

4.2.5.2 Configure an OSA express device

Select an OSA Express device from the list of network devices and specify a port number. Enter the values for the read, write and data channels. Choose whether to enable OSI Layer 2 support.

The port number is required for the new 2 port OSA Express 3 Network devices. If you are not using an OSA Express 3 device, enter 0. OSA Express cards can also run in the OSI layer 2 support mode or the older more common layer 3 mode. The card mode affects all systems that share the device, including systems on other LPARs. If in doubt, specify 2 for compatibility with the default mode used by other operating systems such as z/VM and z/OS. Consult with your hardware administrator for further information on these options.

Example 4.6: Network device driver parameters
Choose the network device.

 1) IBM parallel CTC Adapter (0.0.0600)
 2) IBM parallel CTC Adapter (0.0.0601)
 3) IBM parallel CTC Adapter (0.0.0602)
 4) IBM Hipersocket (0.0.0800)
 5) IBM Hipersocket (0.0.0801)
 6) IBM Hipersocket (0.0.0802)
 7) IBM OSA Express Network card (0.0.0700)
 8) IBM OSA Express Network card (0.0.0701)
 9) IBM OSA Express Network card (0.0.0702)
10) IBM OSA Express Network card (0.0.f400)
11) IBM OSA Express Network card (0.0.f401)
12) IBM OSA Express Network card (0.0.f402)
13) IBM IUCV

> 7

Enter the relative port number. (Enter '+++' to abort).
> 0

Device address for read channel. (Enter '+++' to abort).
[0.0.0700]> 0.0.0700

Device address for write channel. (Enter '+++' to abort).
[0.0.0701]> 0.0.0701

Device address for data channel. (Enter '+++' to abort).
[0.0.0702]> 0.0.0702

Enable OSI Layer 2 support?

0) <-- Back <--
1) Yes
2) No

> 1

MAC address. (Enter '+++' to abort).
> +++

4.2.5.3 Set up the network and select the installation source

After all network device parameters have been entered, the respective driver is installed and you see the corresponding kernel messages.

Next, you need to specify whether to use DHCP autoconfiguration for setting up the network interface parameters. Because DHCP only works on a few devices and requires special hardware configuration settings, choose NO. Doing this prompts you to specify the following networking parameters:

  • The IP address of the system to install

  • The corresponding netmask (if not having been specified with the IP address)

  • The IP address of a gateway to reach the server

  • A list of search domains covered by the domain name server (DNS)

  • The IP address of your domain name server

Example 4.7: Networking parameters
Automatic configuration via DHCP?

0) <-- Back <--
1) Yes
2) No

> 2

Enter your IP address with network prefix.

You can enter more than one, separated by space, if necessary.
Leave empty for autoconfig.

Examples: 192.168.5.77/24 2001:db8:75:fff::3/64. (Enter '+++' to abort).
> 192.168.0.20/24

Enter your name server IP address.

You can enter more than one, separated by space, if necessary.
Leave empty if you don't need one.

Examples: 192.168.5.77 2001:db8:75:fff::3. (Enter '+++' to abort).
> 192.168.0.1

Enter your search domains, separated by a space:. (Enter '+++' to abort).
> example.com

Enter the IP address of your name server. Leave empty if you do not need one. (En
ter '+++' to abort).
> 192.168.0.1

Finally, provide the required information about the installation server, such as the IP address, the directory containing the installation data, and login credentials. The installation system loads when the required information has been provided.

4.2.6 Connecting to the SUSE Linux Enterprise Micro installation system

After loading the installation system, linuxrc prompts you to choose what type of display to use to control the installation procedure. The available options include Remote X11 (X Window System), VNC (Virtual Network Computing protocol), SSH (text mode or X11 installation via Secure Shell), Text-based UI and Graphical UI. The latter starts YaST in graphical mode on a local graphics display if it exists. On the s390x architecture, a local graphics display can be implemented using QEMU and the virtio-gpu driver.

The recommended options are VNC or SSH.

If the Text-based UI option is selected, YaST starts in text mode, and you can perform the installation directly within your terminal. The Text-based UI option is only useful when installing into LPAR.

Note
Note: Terminal emulation for Text-based UI

To be able to work with YaST in the text mode, it needs to run in a terminal with VT220/Linux emulation (also called Text-based UI).

4.2.6.1 Initiating the installation for VNC

To remotely control an installation via VNC, follow these steps:

  1. Selecting the VNC option starts the VNC server. A short note in the console displays the IP address and display number for connecting with vncviewer.

  2. Enter the IP address and the display number of the SUSE Linux Enterprise Micro installation system when prompted to do so.

  3. When prompted, enter the IP address and the display number of the SUSE Linux Enterprise Micro installation system.

    http://<IP address of installation system>:5801/
  4. After the connection has been established, install SUSE Linux Enterprise Micro with YaST.

4.2.6.2 Initiating the installation for the X Window system

Important
Important: X authentication mechanism

The direct installation with the X Window System relies on an authentication mechanism based on host names. This mechanism is disabled in current SUSE Linux Enterprise Micro versions. We recommend performing the installation using SSH or VNC.

To remotely control an installation via X forwarding, follow these steps:

  1. Make sure that the X server allows the client (the system that is installed) to connect. Set the variable DISPLAYMANAGER_XSERVER_TCP_PORT_6000_OPEN="yes" in the file /etc/sysconfig/displaymanager. Restart the X server and allow client binding to the server using xhost CLIENT_IP_ADDRESS.

  2. When prompted at the installation system, enter the IP address of the machine running the X server.

  3. Wait until YaST opens, then start the installation.

4.2.6.3 Initiating the installation for SSH

To connect to an installation system with the name earth via SSH, use the ssh -X earth command. If your workstation runs on Microsoft Windows, use the Putty tool available from http://www.chiark.greenend.org.uk/~sgtatham/putty/. Set Enable X11 forwarding in Putty under Connection › SSH › X11.

If you use another operating system, execute ssh -X earth to connect to an installation system with the name earth. X-Forwarding over SSH is supported if you have a local X server available. Otherwise, YaST provides a text interface over ncurses.

When prompted, enter the root user name and log in with your password. Enter yast.ssh to start YaST. YaST then guides you through the installation.

Important
Important: Fixing YaST over SSH issue

In certain situations, running the GUI version of YaST over SSH with X forwarding may fail with the following error message:

XIO: fatal IO error 11 (Resource temporarily unavailable) on X server "localhost:11.0"

In this case you have two options.

  • Run YaST with the QT_XCB_GL_INTEGRATION=none option, for example:

    QT_XCB_GL_INTEGRATION=none yast.ssh
    QT_XCB_GL_INTEGRATION=none yast2 disk
  • Run the ncurses version of YaST application by disabling X forwarding or by specifying ncurses as the desired UI. To do the latter, use the yast2 disk --ncurses or YUI_PREFERED_BACKEND=ncurses yast2 disk command.

Proceed with the installation procedure as described in Chapter 12, Installation steps.

4.2.7 The SUSE Linux Enterprise Micro boot procedure on IBM Z

On SLES 10 and 11 the boot process was handled by the zipl boot loader. To enable booting from Btrfs partitions and supporting system rollbacks with Snapper, the way SUSE Linux Enterprise Micro is booted on IBM Z has changed.

GRUB 2 replaces zipl on SUSE Linux Enterprise Micro for IBM Z. GRUB 2 on the AMD64/Intel 64 architecture includes device drivers on the firmware level to access the file system. On the mainframe there is no firmware and adding ccw to GRUB 2 would not only be a major undertaking, but would also require a reimplementation of zipl in GRUB 2. Therefore SUSE Linux Enterprise Micro uses a two-stage approach:

Stage one:

A separate partition containing the kernel and an initrd is mounted to /boot/zipl. This kernel and the initrd are loaded via zipl using the configuration from /boot/zipl/config.

This configuration adds the keyword initgrub to the kernel command line. When the kernel and initrd are loaded, the initrd activates the devices required to mount the root file system (see /boot/zipl/active_devices.txt). Afterward a GRUB 2 user space program is started, which reads /boot/grub2/grub.cfg.

Stage two:

The kernel and the initrd specified in /boot/grub2/grub.cfg are started via kexec. Devices listed in /boot/zipl/active_devices.txt that are necessary for starting the on-disk system are then activated. Other devices from that list will be whitelisted, but otherwise ignored. The root file system is mounted and the boot procedure continues like on the other architectures.

4.3 Secure boot

For the secure boot functionality to work on a IBM Z system, the following conditions must be met.

  • The machine must be z15 T01, z15 T02, LinuxONE III LT1, LinuxONE III LT2, or a later model.

  • You must use an LPAR (secure boot is not supported on z/VM and KVM).

  • The LPAR must have secure boot enabled.

  • You must use SCSI (FCP) disks (secure boot is not supported on DASD).

Note
Note: Hardware migration

In case you migrate to a different machine (for example, from z13 to z15), ensure that the LPAR on the target machine has the secure boot state of the system on its disk.

Changing the secure boot state must be performed according to the following procedure.

Procedure 4.4: Changing secure boot state
  1. Enable secure boot in YaST and write the new boot loader.

  2. Shut down the system.

  3. Change the configuration of the LPAR (enable or disable secure boot).

  4. Boot the system.

Note
Note: Secure boot on HMC

The system on the disk configured with the secure=1 parameter can be booted on z15 HMC as long as the firmware supports the new on-disk format (which is always the case on z15).

4.4 The parmfile—automating the system configuration

The installation process can be partially automated by specifying the essential parameters in the parmfile. The parmfile contains all the data required for network setup and DASD configuration. In addition to that, it can be used to set up the connection method to the SUSE Linux Enterprise Micro installation system and the YaST instance running there. This reduces user interaction to the actual YaST installation.

The parameters listed in Section 4.4.1, “General parameters” can be passed to the installation routine as the default values for installation. Note that all IP addresses, server names, and numerical values are examples. Replace them with the actual values of your installation scenario.

The number of lines in the parmfile is limited to 10. You can specify more than one parameter on a line. Parameter names are not case-sensitive. Parameters must be separated by spaces. You may specify the parameters in any order. Always keep the PARAMETER=value string together on one line. The length of each line must not exceed 80 characters. For example:

Hostname=s390zvm01.suse.de HostIP=10.11.134.65
Tip
Tip: Using IPv6 during the installation

By default, you can only assign IPv4 network addresses to your machine. To enable IPv6 during installation, specify one of the following parameters at the boot prompt: ipv6=1 (accept IPv4 and IPv6) or ipv6only=1 (accept IPv6 only).

Some of the following parameters are required. If they are missing, the automatic process prompts you to specify them.

4.4.1 General parameters

AutoYaST=<URL> Manual=0

The AutoYaST parameter specifies the location of the autoinst.xml control file for automatic installation. The Manual parameter controls if the other parameters are only default values that still must be acknowledged by the user. Set this parameter to 0 if all values should be accepted and no questions asked. Setting AutoYaST defaults Manual to 0.

DeviceAutoConfig=<0|1|2>

In linuxrc, the DeviceAutoConfig parameter controls the use of I/O device auto-configuration data for IBM Z systems.

If set to 0, auto-configuration is disabled. If set to 1, the existing auto-config data are applied. If set to 2 (the default), a dialog is shown if auto-config data are present. The user is asked whether to apply them.

For more details, see Section 4.4.4, “I/O device auto-configuration on IBM Z systems”.

Info=<URL>

Specifies a location for a file with additional options. This helps to overcome the limitations of 10 lines (and 80 characters per line under z/VM) for the parmfile. Further documentation on the Info file can be found in Section 9.3.3, “Combining the linuxrc info file with the AutoYaST control file”. Since the Info file can typically only be accessed through the network on IBM Z, you cannot use it to specify the options required to set up the network (that is, the options described in Section 4.4.2, “Configuring the network interface”). Other linuxrc-specific options, such as those related to debugging, must be specified in the parmfile itself.

Upgrade=<0|1>

To upgrade SUSE Linux Enterprise, specify Upgrade=1. A custom parmfile is required for upgrading an existing installation of SUSE Linux Enterprise. Without this parameter, the installation provides no upgrade option.

4.4.2 Configuring the network interface

Important
Important: Configuring the network interface

The settings described in this section apply only to the network interface used during installation.

Hostname=zsystems.example.com

Enter the fully qualified host name.

Domain=example.com

Domain search path for DNS. Allows you to use short host names instead of fully qualified ones.

HostIP=192.168.1.2/24

Enter the IP address of the interface to configure.

Gateway=192.168.1.3

Specify the gateway to use.

Nameserver=192.168.1.4

Specify the DNS server in charge.

InstNetDev=osa

Enter the type of interface to configure. Possible values are osa, hsi, ctc, escon, and iucv (CTC, ESCON, and IUCV are no longer supported).

For the ctc interfaces escon and iucv (CTC, ESCON, and IUCV are no longer supported), enter the IP address of the peer:

Pointopoint=192.168.55.20
OsaInterface=<lcs|qdio>

For osa network devices, specify the host interface (qdio or lcs).

Layer2=<0|1>

For osa QDIO Ethernet and hsi devices, specify whether to enable (1) or disable (0) OSI Layer 2 support.

OSAHWAddr=02:00:65:00:01:09

For Layer 2-enabled osa QDIO Ethernet devices. Either specify a MAC address manually or state OSAHWADDR= (with trailing white space) for the system default.

PortNo=<0|1>

For osa network devices, specify the port number (provided the device supports this feature). The default value is 0.

Each of the interfaces requires certain setup options:

  • Interfaces ctc and escon (CTC and ESCON are no longer supported):

    ReadChannel=0.0.0600
    WriteChannel=0.0.0601

    ReadChannel specifies the READ channel to use. WriteChannel specifies the WRITE channel.

  • For the ctc interface (no longer supported), specify the protocol that should be used for this interface:

    CTCProtocol=<0/1/2>

    Valid entries would be:

    0

    Compatibility mode, also for non-Linux peers other than OS/390 and z/OS (this is the default mode)

    1

    Extended mode

    2

    Compatibility mode with OS/390 and z/OS

  • Network device type osa with interface lcs:

    ReadChannel=0.0.0124

    ReadChannel stands for the channel number used in this setup. A second port number can be derived from this by adding one to ReadChannel. Portnumber is used to specify the relative port.

  • Interface iucv:

    IUCVPeer=PEER

    Enter the name of the peer machine.

  • Network device type osa with interface qdio for OSA-Express Gigabit Ethernet:

    ReadChannel=0.0.0700
    WriteChannel=0.0.0701
    DataChannel=0.0.0702

    For ReadChannel, enter the number of the READ channel. For WriteChannel, enter the number of the WRITE channel. DataChannel specifies the DATA channel. Make sure that the READ channel has an even device number.

  • Interface hsi for HiperSockets and VM guest LANs:

    ReadChannel=0.0.0800
    WriteChannel=0.0.0801
    DataChannel=0.0.0802

    For ReadChannel, enter the appropriate number for the READ channel. For WriteChannel and DataChannel, enter the WRITE and DATA channel numbers.

4.4.3 Specifying the installation source and YaST interface

Install=nfs://server/directory/DVD1/

Specify the location of the installation source to use. Supported protocols are nfs, smb (Samba/CIFS), ftp, tftp http, and https.

If an ftp, tftp or smb URL is provided, specify the user name and password. Skip credentials for anonymous or guest login.

Install=ftp://USER:PASSWORD@SERVER/DIRECTORY/DVD1/
Install=tftp://USER:PASSWORD@SERVER/DIRECTORY/DVD1/

If you want to perform the installation over an encrypted connection, use an https URL. If the certificate cannot be verified, use the sslcerts=0 boot option to disable certificate checking.

In case of a Samba or CIFS installation, you can also specify the domain:

Install=smb://WORKDOMAIN;USER:PASSWORD@SERVER/DIRECTORY/DVD1/
ssh=1 vnc=1 Display_IP=192.168.42.42

The installation method depends on which parameter you specify. ssh enables SSH installation, vnc starts a VNC server on the installing machine, and Display_IP causes the installing system to try to connect to an X server at the specified address. Only one of these parameters should be set.

Important
Important: X authentication mechanism

The direct installation with the X Window System relies on an authentication mechanism based on host names. This mechanism is disabled on current SUSE Linux Enterprise Micro versions. We recommend to perform an installation using SSH or VNC is preferred.

To allow a connection between YaST and the remote X server, run xhost <IP address> with the address of the installing machine on the remote machine.

For VNC, specify a password of six to eight characters to use for installation:

VNCPassword=<a password>

For SSH, specify a password of six to eight characters to use for installation:

ssh.password=<a password>

4.4.4 I/O device auto-configuration on IBM Z systems

I/O device auto-configuration is a mechanism that allows users to specify IDs and settings of I/O devices that should be automatically enabled in Linux. This information is specified for an LPAR via an HMC running in DPM (Dynamic Partition Manager) mode.

Note
Note

The I/O device auto-configuration functionality is available on systems with the DPM running. DPM runs by default on LinuxONE machines. For IBM Z, this functionality must be ordered.

In linuxrc, the DeviceAutoConfig parameter controls the use of I/O device auto-configuration data for IBM Z systems.

DeviceAutoConfig=0

If set to 0, auto-configuration is disabled.

DeviceAutoConfig=1

If set to 1, existing auto-config data are applied.

DeviceAutoConfig=2 (default)

If set to 2 (the default), a dialog is shown if auto-config data are present. The user is asked whether to apply them.

If device auto-config is disabled by the user, the kernel parameter rd.zdev=no-auto is added to the boot options of the target system.

To enable I/O auto-configuration using YaST, run the yast2 system_settings command, switch to the Kernel Settings section, and enable the Enable I/O device auto-configuration option.

To disable I/O auto-configuration in an AutoYaST profile, add the following kernel parameter in the append section of the global boot loader options. For example:

<bootloader>
  <global>
    <append>rd.zdev=no-auto</append>
  </global>
</bootloader>

For more context on the AutoYaST boot loader options, see Section 4.4, “The boot loader”.

During installation, the status of the auto-configuration setting is displayed in the Device Settings section of the Installation Settings screen.

4.4.5 Example parmfiles

The maximum capacity of a parmfile is 860 characters. As a rule of thumb, the parmfile should contain a maximum of 10 lines with no more than 79 characters. When reading a parmfile, all lines are concatenated without adding white spaces, therefore the last character (79) of each line needs to be a Space.

To receive potential error messages on the console, use

linuxrclog=/dev/console
Example 4.8: Parmfile for an installation from NFS with VNC and AutoYaST, with I/O device auto configuration
ramdisk_size=131072 root=/dev/ram1 ro init=/linuxrc TERM=dumb
instnetdev=osa osainterface=qdio layer2=1 osahwaddr=
pointopoint=192.168.0.1 hostip=192.168.0.2
nameserver=192.168.0.3 DeviceAutoConfig=1
install=nfs://192.168.0.4/SLES/SLES-12-Server/s390x/DVD1
autoyast=http://192.168.0.5/autoinst.xml
linuxrclog=/dev/console vnc=1 VNCPassword=testing
Example 4.9: Parmfile for installation with NFS, SSH, and HSI and AutoYaST with NFS
ramdisk_size=131072 root=/dev/ram1 ro init=/linuxrc TERM=dumb
AutoYast=nfs://192.168.1.1/autoinst/s390.xml
Hostname=zsystems.example.com HostIP=192.168.1.2
Gateway=192.168.1.3 Nameserver=192.168.1.4
InstNetDev=hsi layer2=0
Netmask=255.255.255.128 Broadcast=192.168.1.255
readchannel=0.0.702c writechannel=0.0.702d datachannel=0.0.702e
install=nfs://192.168.1.5/SLES-12-Server/s390x/DVD1/
ssh=1 ssh.password=testing linuxrclog=/dev/console
Example 4.10: Parmfile for installation in VLAN
ro ramdisk_size=50000 MANUAL=0 PORTNO=1 ReadChannel=0.0.b140
WriteChannel=0.0.b141 DataChannel=0.0.b142
cio_ignore=all,!condev,!0.0.b140-0.0.b142,!0.0.e92c,!0.0.5000,!0.0.5040
HostIP= Gateway= Hostname=zsystems.example.com nameserver=192.168.0.1
Install=ftp://user:password@10.0.0.1/s390x/SLES15.0/INST/ usevnc=1
vncpassword=12345 InstNetDev=osa Layer2=1 OSAInterface=qdio ssl_certs=0
osahwaddr= domain=example.com self_update=0
vlanid=201

4.5 Using the vt220 terminal emulator

Recent MicroCode Levels allow the use of an integrated vt220 terminal emulator (ASCII terminal) in addition to the standard line mode terminal. The vt220 terminal is connected to /dev/ttysclp0. The line mode terminal is connected to /dev/ttysclp_line0. For LPAR installations, the vt220 terminal emulator is activated by default.

To start the Text-based UI on HMC, log in to the HMC, and select Systems Management › Systems › IMAGE_ID . Select the radio button for the LPAR and select Recovery › Integrated ASCII Console.

To redirect the kernel messages at boot time from the system console to the vt220 terminal, add the following entries to the parameters line in /etc/zipl.conf:

console=ttysclp0 console=ttysclp_line0

The resulting parameters line would look like the following example:

parameters = "root=/dev/dasda2 TERM=dumb console=ttysclp0 console=ttysclp_line0"

Save the changes in /etc/zipl.conf, run zipl, and reboot the system.

4.6 More information

Find further technical documentation about IBM Z in the IBM Redbooks (https://www.redbooks.ibm.com/Redbooks.nsf/domains/zsystems) or at IBM developerWorks (https://developer.ibm.com/). SUSE Linux Enterprise Micro-specific documentation is available from https://developer.ibm.com/technologies/linux/.

4.6.1 General documents about Linux on IBM Z

A general coverage of Linux on IBM Z can be found in the following documents:

  • Linux on IBM eServer zSeries and S/390: ISP and ASP Solutions (SG24-6299)

These documents might not reflect the current state of Linux, but the principles of Linux deployment outlined there remain accurate.

4.6.2 Technical issues of Linux on IBM Z

Refer to the following documents for technical information about the Linux kernel and application topics. For the most recent versions of the documents, visit (https://developer.ibm.com/technologies/linux/).

  • Linux on System z Device Drivers, Features, and Commands

  • zSeries ELF Application Binary Interface Supplement

  • Linux on System z Device Drivers, Using the Dump Tools

  • IBM zEnterprise 196 Technical Guide

  • IBM zEnterprise EC12 Technical Guide

  • IBM z13 Technical Guide

  • IBM z14 Technical Guide

  • IBM z15 Technical Guide

A Redbook for Linux application development is available at http://www.redbooks.ibm.com:

  • Linux on IBM eServer zSeries and S/390: Application Development (SG24-6807)

4.6.3 Advanced configurations for Linux on IBM Z

Refer to the following Redbooks, Redpapers, and online resources for more complex IBM Z scenarios: