--- /dev/null
+.TH lconf 1 "2003 Oct 8" Lustre "configuration utilities"
+.SH NAME
+lconf \- Lustre filesystem configuration utility
+.SH SYNOPSIS
+.br
+.B lconf
+[--node <node_name>] [-d,--cleanup] [--noexec] [--gdb] [--nosetup] [--nomod] [-n,--noexec] [-v,--verbose] [-h,--help] <XML-config file>
+[options] --add <objecttype> [args]
+.br
+.SH DESCRIPTION
+.B lconf
+, when invoked configures a node following directives in the <XML-config-file>. There will be single configuration file for all the nodes in a single cluster. This file should be distributed to all the nodes in the cluster or kept in a location accessible to all the nodes. One option is to store the cluster configuration information in LDAP format on an LDAP server that can be reached from all the cluster nodes.
+.PP
+The arguments that can be used for lconf are:
+.PP
+.TP
+--config <arg>
+Cluster configuration name used for LDAP query
+.TP
+--d|--cleanup
+Unconfigure a node. The same config and --node argument used for configuration needs to be used for cleanup as well. This will attempt to undo all of the configuration steps done by lconf, including unloading the kernel modules.
+.TP
+--dump <file>
+Dump the kernel debug log to the specified file before portals is unloaded during cleanup.
+.TP
+--dump_path <arg>
+Path to save debug dumps. Default is /tmp/lustre_log
+.TP
+--failover
+Used to shutdown without saving state. Default is 0. This will allow the node to give up service to another node for failover purposes. This will not be a clean shutdown.
+.TP
+--force
+Forced unmounting and/or obd detach during cleanup. Default is 0.
+.TP
+--gdb
+Causes lconf to print a message and pause for 5 seconds after creating a gdb module script and before doing any Lustre configuration (the gdb module script is always created, however).
+.TP
+--gdb_script <arg>
+Full name of gdb debug script. Default is /tmp/ogdb.
+.TP
+--group <arg>
+The group of devices to cleanup/configure.
+.TP
+-h,--help
+Print help.
+.TP
+--inactive <UUID>
+The UUID of the service to be ignored by a client mounting Lustre. Allows the client to mount in the presence of some inactive services. (currently OST only). Multiple UUIDs can be specified by repeating the option.
+.TP
+--lctl-dump <arg>
+Dump all ioctls to the specified file
+.TP
+--ldapurl <arg>
+LDAP server URL
+.TP
+--lustre_upcall <path>
+Set the location of the Lustre upcall scripts used by the client for recovery
+.TP
+--lustre=src_dir
+Specify the base directory for Lustre sources, this parameter will cause lconf to load the lustre modules from this soure tree.
+.TP
+--mds_ost_conn
+Open connections to OSTs on MDS.
+.TP
+--maxlevel <level>
+Perform configuration of devices and services up to level given. level can take the values net, dev, svc, fs. When used in conjunction with cleanup, services are torn down up to a certain level. Default is 100.
+.TP
+--minlevel <level>
+Specify the minimum level of services to configure/cleanup. Default is 0.
+.TP
+--node node_name
+Specify a specific node to configure. By default, lconf will search for nodes with the local hostname and 'localhost'. When --node is used, only node_name is searched for. If a matching node is not found in the config, then lconf exits with an error.
+.TP
+--noexec,-n
+Print, but don't execute, the steps lconf will perform. This is useful for debugging a configuration, and when used with --node, can be run on any host.
+.TP
+--nomod
+Only setup devices and services, do not load modules.
+.TP
+--nosetup
+Only load modules, do not configure devices or services.
+.TP
+--portals_upcall <path>
+Specify the location of the Portals upcall scripts used by the client for recovery
+.TP
+--ptldebug debug-level
+This options can be used to set the required debug level.
+.TP
+--recover <arg>
+Recover a device.
+.TP
+--reformat
+Reformat all the devices. This is essential on the first time the file system is brought up.
+.TP
+--select <arg>
+Select a particular node for a service
+.TP
+--subsystem <arg>
+Set the portals debug subsystem.
+.TP
+--timeout <arg>
+Set the recovery timeout period.
+.TP
+--upcall <path>
+Set the location of both Lustre and Portals upcall scripts used by the client for recovery
+.TP
+--verbose,-v
+Be verbose and show actions while going along.
+.TP
+--write_conf
+Save all client configuration information on the MDS
+.SH EXAMPLES
+.TP
+.B lconf --node client config.xml
+This invokes lconf on the client node.
+.TP
+.B lconf --ptldebug "~(portals | malloc | trace)"
+Used to set the required debug levels (all but these).
+.TP
+.B lconf --ptldebug "ldlm|ha"
+Used to turn-on specific debug types.
+.TP
+.B lconf --inactive OST_ost1_UUID --inactive OST_ost2_UUID config.xml
+A subset of failed OSTs can be ignored during Lustre mount on the clients by using this option. Here OST1 and OST2 have failed and need to be ignored.
+.SH BUGS
+None are known.
--- /dev/null
+.TH lmc 1 "2003 Oct 8" Lustre "configuration utilities"
+.SH NAME
+lmc \- Lustre configuration maker
+.SH SYNOPSIS
+.br
+.B lmc
+[options] --add <objecttype> [args]
+.br
+.SH DESCRIPTION
+.B lmc
+, when invoked, adds configuration data to a configuration file. In future,
+lmc will also be able to remove configuration data or convert its format.
+A Lustre cluster consists of several components - MDSs, client mount-points,
+OSTs, LOVs and networks. A single configuration file would be generated for the complete cluster. In the lmc command line interface, each of these components is associated with an objecttype.
+.PP
+The objecttype refers to a collection of related configuration entities and can be one of
+.B net
+,
+.B MDS
+,
+.B LOV
+,
+.B OST
+,
+.B mtpt
+,
+.B route
+or
+.B echo-client.
+.PP
+We describe the arguments required for the addition of each objecttype.
+.PP
+To generate configuration data associated with systems in a Lustre cluster:
+.PP
+.B --add node
+Adds a new node in the cluster configuration.
+The arguements required are:
+.TP
+--node <node_name>
+This will create a new node with the given name if not already present.
+.TP
+--timeout <num>
+Timeout before going into recovery
+.TP
+--lustre_upcall <path>
+Set the location of the Lustre upcall scripts used by the client for recovery
+.TP
+--portals_upcall <path>
+Specify the location of the Portals upcall scripts used by the client for recovery
+.TP
+--upcall <path>
+Specify the location of both (Lustre and Portals) upcall scripts used by the client for recovery
+.PP
+.B --add net
+Adds a network device descriptor for the given node, with parameters as indicated.
+The arguments required are:
+.TP 12
+--node <node_name>
+This will create a new node with the given name if not already present. This is also used to specify a specific node for other elements.
+.TP
+--nettype <type>
+This can be tcp, elan, gm, scimac.
+.TP
+--nid nid
+The network id, e.g. ElanID or IP address as used by Portals. If nid is '*', then the local address of the interface with specified nettype is will be substituted when the node is configured with lconf. An nid of '*' should be used only for the generic client configuration.
+.TP
+--hostaddr addr
+.TP
+--router
+Optional flag to mark this node as a router
+.TP
+--port [port]
+Optional arguement to indicate the tcp port. The default is 988.
+.TP
+--tcpbuf <size>
+Optional arguement. The default TCP buffer size is 1MB.
+.TP
+--irq_affinity 0|1
+Optional arguement. Default is 0.
+.TP
+--nid_exchange 0|1
+Optional arguement since some OSTs might not have the required support. This is turned off by default, value of 1 will turn it ON.
+.PP
+.B --add mds
+Specify the MDS configuration
+.TP
+--node <node name>
+Name of the node on which the MDS resides
+.TP
+--mds <mds_name>
+Common name of the MDS
+.TP
+--dev <pathname>
+Path of device on local system. If the is a file, then a loop device is created and used as the block device.
+.TP
+--size <size>
+Optional argument indicating the size (in KB) of the device to be created (used typically for loop devices).
+.TP
+--node <nodename>
+Adds an MDS to the specified node. This requires a --node argument, and it must not be a profile node.
+.TP
+--fstype extN|ext3
+Optional argument used to specify the file system type. Default is ext3.
+.TP
+--journal_size <size>
+Optional arguement to specify the journal size for the ext2/ext3 file system. The size should be in the units expected by mkfs, so for ext3 it should be in MB. If this is option is not used, the ext2/ext3 filesystem will be configured with the default journal size.
+.PP
+.B --add lov
+Creates an LOV with the specified parameters. The mds_name must already exist in the descriptor.
+.TP
+--lov <name>
+Common name for the LOV
+.TP
+--mds <name>
+Common name for the MDS
+.TP
+--stripe_sz <size>
+Stripe size
+.TP
+--stripe_cnt <count>
+A value of 0 for this means to stripe on all available OSTs. Default is 0.
+.TP
+--stripe_pattern <pattern>
+Only Pattern 0 (RAID 0) is supported currently.
+.PP
+.B --add ost
+Creates an OBD, OST, and OSC. The OST and OBD are created on the specified node.
+.TP
+--ost <name>
+Assign a name to the OST device.
+.TP
+--node <nodename>
+Node on which the OST service is run, can not be a profile node.
+.TP
+--dev <pathname>
+Path of device on local system. If this is a file, then a loop device is created and used as the block device.
+.TP
+--size [size]
+Optional argument indicating the size (in KB) of the device to be created (used typically for loop devices).
+.TP
+--obdtype
+obdfilter|obdecho
+.TP
+--lov <name>
+Optional arguement. Name of LOV to which this OSC will be attached.
+.TP
+--ostuuid UUID
+Specify the UUID of the OST device.
+.TP
+--fstype
+extN|ext3 Optional arguement used to specify the file system type. Default is ext3.
+.TP
+--journal_size <size>
+Optional arguement to specify the journal size for the ext2/ext3 file system. The size should be in the units expected by mkfs, so for ext3 it should be in MB. If this is option is not used, the ext2/ext3 filesystem will be configured with the default journal size.
+.PP
+.B --add mtpt
+Creates a mount-point on the specified node. Either an LOV or OSC name can be used.
+.TP
+--node node
+Node that will use the mtpt.
+.TP
+--path /mnt/path
+The mount-point to use to mount Lustre filesystem
+.TP
+--mds mds_name
+MDS name
+.TP
+--ost ost_name | --lov lov_name
+OST or LOV name as specified earlier in the configuration
+.PP
+.B --add route
+Creates a static route through a gateway to a specific nid or a range of nid's.
+.TP
+--node node
+Node to add the route to.
+.TP
+--gw nid
+The nid of the gateway (must be a local interface or a peer).
+.TP
+--tgt nid
+For a specific route, this is the target nid.
+.TP
+--lo nid
+For a range route, this is the lo value nid.
+.TP
+--hi nid
+For a range route, this is the hi value nid.
+.PP
+.B --add echo-client
+Used for testing purpose only.
+.TP
+--node node
+Name of the node that echo client should run on,
+.TP
+--obd obd_name
+.SH OPTIONS
+One of the following options should be specified.
+.TP 12
+--output filename
+Send output to the file. If the file exists, it will be overwritten.
+.TP
+--merge filename
+Add the new element to an existing file.
+.SH EXAMPLES
+Real life examples are given in the Lustre-conf manual page.
+.SH BUGS
+None are known.
git://git.whamcloud.com / fs / lustre-release.git / commitdiff