node - a logical object definition in the xCAT database.
node Attributes: addkcmdline, appstatus, appstatustime, arch, bmc, bmcpassword, bmcport, bmcusername, chain, chassis, cons, conserver, currchain, currstate, getmac, groups, hcp, hidden, hostnames, hwtype, id, initrd, installnic, interface, ip, iscsipassword, iscsiserver, iscsitarget, iscsiuserid, kcmdline, kernel, mac, mgt, migrationdest, monserver, mpa, mtm, netboot, nfsdir, nfsserver, nimserver, node, nodetype, ondiscover, os, otherinterfaces, parent, passwd.HMC, passwd.admin, passwd.celogin, passwd.general, passwd.hscroot, password, postbootscripts, postscripts, power, pprofile, prescripts-begin, prescripts-end, primarynic, primarysn, profile, provmethod, rack, room, routenames, serial, serialflow, serialport, serialspeed, servicenode, setupconserver, setupdhcp, setupftp, setupipforward, setupldap, setupnameserver, setupnfs, setupnim, setupntp, setuptftp, sfp, side, slot, status, statustime, supernode, supportedarchs, switch, switchinterface, switchport, switchvlan, termport, termserver, tftpserver, unit, usercomment, userid, username, vmbeacon, vmbootorder, vmcfgstore, vmcpus, vmhost, vmmemory, vmnicnicmodel, vmnics, vmstorage, vmstoragemodel, vmtextconsole, vmvirtflags, vmvncport, webport, xcatmaster
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
User specified one or more parameters to be passed to the kernel
A comma-delimited list of application status. For example: 'sshd=up,ftp=down,ll=down'
The date and time when appstatus was updated.
The hardware architecture of this node. Valid values: x86_64, ppc64, x86, ia64.
The hostname of the BMC adapater.
The BMC password. If not specified, the key=ipmi row in the passwd table is used as the default.
In systems with selectable shared/dedicated ethernet ports, this parameter can be used to specify the preferred port. 0 means use the shared port, 1 means dedicated, blank is to not assign
The BMC userid. If not specified, the key=ipmi row in the passwd table is used as the default.
A comma-delimited chain of actions to be performed automatically for this node. Valid values: discover, boot or reboot, install or netboot, runcmd=<cmd>, runimage=<image>, shell, standby. (Default - same as no chain). Example, for BMC machines use: runcmd=bmcsetup,standby.
The BladeCenter chassis the blade is in.
The console method. If nodehm.serialport is set, this will default to the nodehm.mgt setting, otherwise it defaults to unused. Valid values: cyclades, mrv, or the values valid for the mgt attribute.
The hostname of the machine where the conserver daemon is running. If not set, the default is the xCAT management node.
The current execution chain for this node. Set by xCAT. Initialized from chain and updated as chain is executed.
The current chain state for this node. Set by xCAT.
The method to use to get MAC address of the node with the getmac command. If not set, the mgt attribute will be used. Valid values: same as values for mgmt attribute.
A comma-delimited list of groups this node is a member of. Group names are arbitrary, except all nodes should be part of the 'all' group.
The hardware control point for this node (HMC, IVM, Frame or CEC). Do not need to set for BPAs and FSPs.
or
The hardware control point for this node.
Used to hide fsp and bpa definitions, 1 means not show them when running lsdef and nodels
Hostname aliases added to /etc/hosts for this node.
The hardware type of the node. Only can be one of fsp, bpa, cec, frame, ivm, hmc and lpar
or
The node type. Valid values: cec (Central Electronic Complex), lpar (logical partition), zvm (z/VM host operating system), and vm (virtual machine).
For LPARs: the LPAR numeric id; for CECs: the cage number; for Frames: the frame number.
or
The slot number of this blade in the BladeCenter chassis.
The initial ramdisk image that network boot actions should use (could be a DOS floppy or hard drive image if using memdisk as kernel)
The network adapter on the node that will be used for OS deployment, the installnic can be set to the network adapter name or the mac address or the keyword ``mac'' which means that the network interface specified by the mac address in the mac table will be used. If not set, primarynic will be used.
The adapter interface name that will be used to install and manage the node. E.g. eth0 (for linux) or en0 (for AIX).)
The IP address of the node.
The password for the iscsi server containing the boot device for this node.
The server containing the iscsi boot device for this node.
The iscsi disk used for the boot device for this node. Filled in by xCAT.
The userid of the iscsi server containing the boot device for this node.
Arguments to be passed to the kernel
The kernel that network boot actions should currently acquire and use. Note this could be a chained boot loader such as memdisk or a non-linux boot loader
The mac address or addresses for which xCAT will manage static bindings for this node. This may be simply a mac address, which would be bound to the node name (such as ``01:02:03:04:05:0E''). This may also be a ``|'' delimited string of ``mac address!hostname'' format (such as ``01:02:03:04:05:0E!node5|01:02:03:05:0F!node6-eth1'').
The method to use to do general hardware management of the node. This attribute is used as the default if power or getmac is not set. Valid values: ipmi, blade, hmc, ivm, fsp. See the power attribute for more details.
A noderange representing candidate destinations for migration (i.e. similar systems, same SAN, or other criteria that xCAT can use
The monitoring aggregation point for this node. The format is ``x,y'' where x is the ip address as known by the management node and y is the ip address as known by the node.
The managment module used to control this blade.
The machine type and model number of the node. E.g. 7984-6BU
The type of network booting to use for this node. Valid values: pxe or xnba for x86* architecture, yaboot for POWER architecture.
The path that should be mounted from the NFS server.
The NFS or HTTP server for this node (as known by this node).
Not used for now. The NIM server for this node (as known by this node).
The hostname of a node in the cluster.
A comma-delimited list of characteristics of this node. Valid values: blade, vm (virtual machine), osi (OS image), mm, rsa, switch.
What to do when a new node is discovered. Valid values: nodediscover.
The operating system deployed on this node. Valid values: AIX, rhels*,rhelc*, rhas*,centos*,SL*, fedora*, sles* (where * is the version #). As a special case, if this is set to ``boottarget'', then it will use the initrd/kernel/parameters specified in the row in the boottarget table in which boottarget.bprofile equals nodetype.profile.
Other IP addresses to add for this node. Format: -<ext>:<ip>,<intfhostname>:<ip>,...
For LPARs: the CEC; for FSPs: the CEC; for CEC: the frame (if one exists); for BPA: the frame; for frame: the building block number (which consists 1 or more service nodes and compute/storage nodes that are serviced by them - optional).
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
Password of the HMC or IVM. If not filled in, xCAT will look in the passwd table for key=hmc or key=ivm. If not in the passwd table, the default used is abc123 for HMCs and padmin for IVMs.
or
Password to use to access the management module. If not specified, the key=blade row in the passwd table is used as the default.
or
Password to use to access the web service.
Comma separated list of scripts that should be run on thisnode after diskfull installation or diskless boot. On AIX these scripts are run during the processing of /etc/inittab. On Linux they are run at the init.d time. xCAT automatically adds the scripts in the xcatdefaults.postbootscripts attri bute to run first in the list.
Comma separated list of scripts that should be run on this node after diskfull installation or diskless boot. xCAT automatically adds the postscripts from the xcatdefaults.postscripts attribute of the table to run first on the nodes after install or diskless boot. For installation of RedHat, CentOS, Fedora, the scripts will be run before the reboot. For installation of SLES, the scripts will be run after the reboot but before the init.d process. For diskless deployment, the scripts will be run at the init.d time, and xCAT will automatically add the list of scripts from the postbootscripts attribute to run after postscripts list. For installation of AIX, the scripts will run after the reboot and acts the same as the postbootscripts attribute. For AIX, use the postbootscripts attribute. Support will be added in the future for the postscripts attribute to run the scripts before the reboot in AIX.
The method to use to control the power of the node. If not set, the mgt attribute will be used. Valid values: ipmi, blade, hmc, ivm, fsp. If ``ipmi'', xCAT will search for this node in the ipmi table for more info. If ``blade'', xCAT will search for this node in the mp table. If ``hmc'', ``ivm'', or ``fsp'', xCAT will search for this node in the ppc table.
The LPAR profile that will be used the next time the LPAR is powered on with rpower. For DFM, the pprofile attribute should be set to blank
The scripts to be run at the beginning of the nodeset(Linux),
nimnodeset(AIX) or mkdsklsnode(AIX) command.
The format is:
[action1:]s1,s2...[|action2:s3,s4,s5...]
where:
- action1 and action2 for Linux are the nodeset actions specified in the command.
For AIX, action1 and action1 can be 'diskless' for mkdsklsnode command'
and 'standalone for nimnodeset command.
- s1 and s2 are the scripts to run for action1 in order.
- s3, s4, and s5 are the scripts to run for actions2.
If actions are omitted, the scripts apply to all actions.
Examples:
myscript1,myscript2 (all actions)
diskless:myscript1,myscript2 (AIX)
install:myscript1,myscript2|netboot:myscript3 (Linux)
All the scripts should be copied to /install/prescripts directory.
The following two environment variables will be passed to each script:
NODES a coma separated list of node names that need to run the script for
ACTION current nodeset action.
If '#xCAT setting:MAX_INSTANCE=number' is specified in the script, the script
will get invoked for each node in parallel, but no more than number of instances
will be invoked at at a time. If it is not specified, the script will be invoked
once for all the nodes.
The scripts to be run at the end of the nodeset(Linux),
nimnodeset(AIX),or mkdsklsnode(AIX) command.
The format is the same as the 'begin' column.
The network adapter on the node that will be used for xCAT management, the primarynic can be set to the network adapter name or the mac address or the keyword ``mac'' which means that the network interface specified by the mac address in the mac table will be used. Default is eth0.
Not used currently. The primary servicenode, used by this node.
The string to use to locate a kickstart or autoyast template to use for OS deployment of this node. If the provmethod attribute is set to an osimage name, that takes precendence. Otherwise, the os, profile, and arch are used to search for the files in /install/custom first, and then in /opt/xcat/share/xcat.
The provisioning method for node deployment. The valid values are install, netboot, statelite or an os image name from the osimage table. If an image name is specified, the osimage definition stored in the osimage table and the linuximage table (for Linux) or nimimage table (for AIX) are used to locate the files for templates, pkglists, syncfiles, etc. On Linux, if install, netboot or statelite is specified, the os, profile, and arch are used to search for the files in /install/custom first, and then in /opt/xcat/share/xcat.
The frame the node is in.
The room the node is in.
A comma separated list of route names that refer to rows in the routes table. These are the routes that should be defined on this node when it is deployed.
The serial number of the node.
The flow control value of the serial port for this node. For SOL this is typically 'hard'.
The serial port for this node, in the linux numbering style (0=COM1/ttyS0, 1=COM2/ttyS1). For SOL on IBM blades, this is typically 1. For rackmount IBM servers, this is typically 0.
The speed of the serial port for this node. For SOL this is typically 19200.
A comma separated list of node names (as known by the management node) that provides most services for this node. The first service node on the list that is accessible will be used. The 2nd node on the list is generally considered to be the backup service node for this node when running commands like snmove.
Do we set up Conserver on this service node? Valid values:yes or 1, no or 0. If yes, configures and starts conserver daemon. If no or 0, it does not change the current state of the service.
Do we set up DHCP on this service node? Not supported on AIX. Valid values:yes or 1, no or 0. If yes, runs makedhcp -n. If no or 0, it does not change the current state of the service.
Do we set up a ftp server on this service node? Not supported on AIX Valid values:yes or 1, no or 0. If yes, configure and start vsftpd. If no or 0, it does not change the current state of the service.
Do we set up ip forwarding on this service node? Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
Do we set up ldap caching proxy on this service node? Not supported on AIX. Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
Do we set up DNS on this service node? Valid values:yes or 1, no or 0. If yes, creates named.conf file with forwarding to the management node and starts named. If no or 0, it does not change the current state of the service.
Do we set up file services (HTTP,FTP,or NFS) on this service node? For AIX will only setup NFS, not HTTP or FTP. Valid values:yes or 1, no or 0.If no or 0, it does not change the current state of the service.
Not used. Do we set up a NIM server on this service node? Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
Not used. Use setupntp postscript to setup a ntp server on this service node? Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
Do we set up TFTP on this service node? Not supported on AIX. Valid values:yes or 1, no or 0. If yes, configures and starts atftp. If no or 0, it does not change the current state of the service.
The Service Focal Point of this Frame. This is the name of the HMC that is responsible for collecting hardware service events for this frame and all of the CECs within this frame.
<BPA>-<port> or <FSP>-<port>. The side information for the BPA/FSP. The side attribute refers to which BPA/FSP, A or B, which is determined by the slot value returned from lsslp command. It also lists the physical port within each BPA/FSP which is determined by the IP address order from the lsslp response. This information is used internally when communicating with the BPAs/FSPs
The slot number of the blade in the chassis.
The current status of this node. This attribute will be set by xCAT software. Valid values: defined, booting, netbooting, booted, discovering, configuring, installing, alive, standingby, powering-off, unreachable. If blank, defined is assumed. The possible status change sequenses are: For installaton: defined->[discovering]->[configuring]->[standingby]->installing->booting->booted->[alive], For diskless deployment: defined->[discovering]->[configuring]->[standingby]->netbooting->booted->[alive], For booting: [alive/unreachable]->booting->[alive], For powering off: [alive]->powering-off->[unreachable], For monitoring: alive->unreachable. Discovering and configuring are for x Series dicovery process. Alive and unreachable are set only when there is a monitoring plug-in start monitor the node status for xCAT. Please note that the status values will not reflect the real node status if you change the state of the node from outside of xCAT (i.e. power off the node using HMC GUI).
The data and time when the status was updated.
Indicates the connectivity of this CEC in the HFI network. A comma separated list of 2 ids. The first one is the supernode number the CEC is part of. The second one is the logical location number (0-3) of this CEC within the supernode.
Comma delimited list of architectures this node can execute.
The switch hostname.
The interface name from the node perspective. This is not currently used by xCAT, but administrators may wish to use this for their own purposes
The port number in the switch that this node is connected to. On a simple 1U switch, an administrator can generally enter the number as printed next to the ports, and xCAT will understand switch representation differences. On stacked switches or switches with line cards, administrators should usually use the CLI representation (i.e. 2/0/1 or 5/8). One notable exception is stacked SMC 8848M switches, in which you must add 56 for the proceeding switch, then the port number. For example, port 3 on the second switch in an SMC8848M stack would be 59
xCAT currently does not make use of this field, however it may do so in the future. For now, it can be used by administrators for their own purposes, but keep in mind some xCAT feature later may try to enforce this if set
The port number on the terminal server that this node is connected to.
The hostname of the terminal server.
The TFTP server for this node (as known by this node).
The vertical position of the node in the frame
Any user-written notes.
The z/VM userID of this node.
Userid of the HMC or IVM. If not filled in, xCAT will look in the passwd table for key=hmc or key=ivm. If not in the passwd table, the default used is hscroot for HMCs and padmin for IVMs.
or
Userid to use to access the management module. If not specified, the key=blade row in the passwd table is used as the default.
or
Userid to use to access the web service.
This flag is used by xCAT to track the state of the identify LED with respect to the VM.
Boot sequence (i.e. net,hd)
Optional location for persistant storage separate of emulated hard drives for virtualization solutions that require persistant store to place configuration data
Number of CPUs the node should see.
The system that currently hosts the VM
Megabytes of memory the VM currently should be set to.
Model of NICs that will be provided to VMs (i.e. e1000, rtl8139, virtio, etc)
Network configuration parameters. Of the general form [physnet:]interface,.. Generally, interface describes the vlan entity (default for native, tagged for tagged, vl[number] for a specific vlan. physnet is a virtual switch name or port description that is used for some virtualization technologies to construct virtual switches. hypervisor.netmap can map names to hypervisor specific layouts, or the descriptions described there may be used directly here where possible.
A list of storage files or devices to be used, pipe delimited. i.e. /cluster/vm/<nodename> for KVM/Xen, or nfs://<server>/path/to/folder/ for VMware
Model of storage devices to provide to guest
Tracks the Psuedo-TTY that maps to the serial port or console of a VM
General flags used by the virtualization method. For example, in Xen it could, among other things, specify paravirtualized setup, or direct kernel boot. For a hypervisor/dom0 entry, it is the virtualization method (i.e. ``xen''). For KVM, the following flag=value pairs are recognized:
imageformat=[raw|fullraw|qcow2]
raw is a generic sparse file that allocates storage on demand
fullraw is a generic, non-sparse file that preallocates all space
qcow2 is a sparse, copy-on-write capable format implemented at the virtualization layer rather than the filesystem level
clonemethod=[qemu-img|reflink]
qemu-img allows use of qcow2 to generate virtualization layer copy-on-write
reflink uses a generic filesystem facility to clone the files on your behalf, but requires filesystem support such as btrfs
Tracks the current VNC display port (currently not meant to be set
The port of the web service.
The hostname of the xCAT service node (as known by this node). This acts as the default value for nfsserver and tftpserver, if they are not set. If xcatmaster is not set, the node will use whoever responds to its boot request as its master. For the directed bootp case for POWER, it will use the management node if xcatmaster is not set.
mkdef(1), chdef(1), lsdef(1), rmdef(1)