1 #LyX 1.3 created this file. For more info see http://www.lyx.org/
10 \papersize letterpaper
15 \use_numerical_citations 0
16 \paperorientation portrait
19 \paragraph_separation skip
21 \quotes_language english
25 \paperpagestyle default
35 lmc Lustre configuration maker.
43 lmc [options] --add <objectype> [args]
48 lmc [options] --remove <objecttype> [args] -- [NOT IMPLEMENTED]
53 lmc [options] --convert [args] -- [NOT IMPLEMENTED]
65 , when invoked, adds configuration data to the config file.
72 will also be able to remove configuration data or convert its format.
73 A Lustre cluster consists of several components - MDS's, mount-points,
74 OSTs, LOVs and whatever those reference (e.g nets and profiles).
75 A single configuration file would be generated for the complete cluster.
82 command line interface, each of these components is associated with an
95 refers to a collection of related configuration entities and can be one
122 echo_client, or cobd.
125 We describe the arguments required for the addition of each
133 To generate configuration data associated with systems in a Lustre cluster:
141 node Adds a new node in the cluster configuration.
146 The arguments required are:
150 ''node_name'' This will create a new node with the given name if not
154 --timeout\SpecialChar ~
155 <num> Timeout before going into recovery
158 --lustre_upcall\SpecialChar ~
159 <path> Set the location of the Lustre upcall scripts used
160 by the client for recovery
163 --portals_upcall\SpecialChar ~
164 <path> Specify the location of the Portals upcall scripts
165 used by the client for recovery
168 --upcall\SpecialChar ~
169 <path> Specify the location of both (Lustre and Portals) upcall
170 scripts used by the client for recovery
173 --group_upcall\SpecialChar ~
174 <path> Specify the location of the group upcall scripts used by the MDS
175 for determiningsupplementary group membership
178 --ptldebug\SpecialChar ~
179 <debug_level> Set the portals debug level.
182 --subsytem\SpecialChar ~
183 <subsystem_name> Specify which Lustre subsystems have debug output
193 net Adds a network device descriptor for the given node, with parameters
198 The arguments required are:
202 ''node_name'' This will create a new node with the given name if not
204 This is also used to specify a specific node for other elements.
207 --nettype\SpecialChar ~
208 <type> Specify the network type.This can be
219 ElanID or IP address as used by Portals.
224 is '*', then the local address of the interface with specified
228 is will be substituted when the node is configured with lconf.
237 should be used only for the generic
244 --cluster_id\SpecialChar ~
245 id Specify the cluster ID.
248 --hostaddr\SpecialChar ~
249 addr Specify the host address.
250 lconf will transfer it to the real host address.
253 --router Optional flag to mark this node as a router
257 [port] Optional arguement to indicate the tcp port.
262 --tcpbuf\SpecialChar ~
263 <size> Optional arguement.
264 The default TCP buffer size is 1MB.
267 --irq_affinity\SpecialChar ~
268 0|1 Optional arguement.
272 --nid_exchange\SpecialChar ~
273 0|1 Optional arguement since some OSTs might not have the
275 This is turned off by default, value of 1 will turn it ON.
287 name> Name of the node on which the MDS resides
291 <mds_name> Specify MDS name.
294 --mdsuuid\SpecialChar ~
295 <uuid> Specify MDS uuid.
298 --failover Enable failover support on MDS.
302 <pathname> Path of device on local system.
303 If the is a file, then a loop device is created and used as the block device.
307 <size> Optional argument indicating the size (in KB) of the device
308 to be created (used typically for loop devices).
312 <nodename> Adds an MDS to the specified node.
317 argument, and it must not be a profile node.
320 --fstype\SpecialChar ~
321 ldiskfs|ext3 Optional argument used to specify the file system type.
322 Default is ext3. For 2.6 kernels the ldiskfs filesystem must be used.
325 --inode_size\SpecialChar ~
326 <size> Specify new inode size for underlying ext3 file system. Must be a
327 power of 2 between 128 and 4096. The default inode size is selected based
328 on the default number of stripes specified for the filesystem.
331 --group_upcall\SpecialChar ~
332 <pathname> The group upcall program to call for resolving a user's secondary
333 groups. The default value is NONE, which means that the MDS will use whatever
334 supplementary group is passed from the client. The supplied upcall is
335 /usr/sbin/l_getgroups, which gets groups from the MDS's /etc/group file
336 based on the client-supplied UID.
338 --mkfsoptions\SpecialChar ~
339 <options> Optional argument to mkfs.
342 --mountfsoptions\SpecialChar ~
343 <options> Optional argument to mount fs.
344 Mount options will be passed by this argument.
345 For example, extents are to be enabled by adding ",extents" to the --mountfsopt
348 \begin_inset Quotes eld
352 \begin_inset Quotes erd
356 \begin_inset Quotes eld
360 \begin_inset Quotes erd
363 can also be added to it.
366 --journal_size\SpecialChar ~
367 <size> Optional arguement to specify the journal size for
368 the ext3 file system.
369 The size should be in the units expected by
373 , so for ext3 it should be in MB.
374 If this is option is not used, the ext3 filesystem will be configured
375 with a journal size dependent upon how large the filesystem is.
380 lov Creates an LOV with the specified parameters.
385 must already exist in the descriptor.
390 <name> Specify LOV name.
394 <name> Specify the mds name.
397 --stripe_sz\SpecialChar ~
398 <size> Specify the stripe size in bytes. This much data is written to each
399 stripe before starting to write in the next stripe. Default is 1048576.
402 --stripe_cnt\SpecialChar ~
403 <count> A value of 0 for this means to use the currently optimal number
404 of stripes. Default is 1 stripe per file.
407 --stripe_pattern\SpecialChar ~
408 <pattern> Only Pattern 0 (RAID 0) is supported currently.
413 ost Creates an OBD, OST, and OSC.
414 The OST and OBD are created on the specified node.
419 <name> Assign a name to the OST device.
423 <nodename> Node on which the OST service is run, can not be a profile
427 --failover Enable failover support on OST.
431 <pathname> Path of device on local system.
432 If this is a file, then a loop device is created and used as the block
437 [size] Optional argument indicating the size (in KB) of the device
438 to be created (used typically for loop devices).
441 --osdtype\SpecialChar ~
442 obdfilter|obdecho Specify the type of OSD.
446 <name> Optional arguement.
447 Name of LOV to which this OSC will be attached.
451 --ostuuid\SpecialChar ~
452 UUID Specify the UUID of the OST device.
456 --fstype\SpecialChar ~
457 ldiskfs|ext3 Optional arguement used to specify the file system type.
461 --inode_size\SpecialChar ~
462 <size> Specify new inode size for underlying ext3 file system.
465 --mkfsoptions\SpecialChar ~
466 <options> Optional argument to mkfs.
469 --mountfsoptions\SpecialChar ~
470 <options> Optional argument to mount fs.
471 Mount options will be passed by this argument.
472 For example, extents are to be enabled by adding ",extents" to the --mountfsopt
475 \begin_inset Quotes eld
479 \begin_inset Quotes erd
483 \begin_inset Quotes eld
487 \begin_inset Quotes erd
490 can also be added to it.
493 --journal_size\SpecialChar ~
494 <size> Optional arguement to specify the journal size for
495 the ext3 file system.
496 The size should be in the units expected by
500 , so for ext3 it should be in MB.
501 If this is option is not used, the ext3 filesystem will be configured
502 with a journal size dependent upon how large the filesystem is.
507 mtpt Creates a mount-point on the specified node for the given LOV.
512 node Node or profile node that will use the
520 /mnt/path Specify the mountpoint for Lustre.
524 mds_name Specify the mds name, which will provide service for this
529 ost_name\SpecialChar ~
532 lov_name Specify the ost or lov, which will provide service
536 --clientoptions Specify the options for Lustre, such as async.
541 route Creates a static route through a gateway to a specific
554 node Node to add the route to.
557 --router Optional flag to mark a node as router.
565 of the gateway (must be a local interface or a peer).
568 --gateway_cluster_id\SpecialChar ~
569 id Specify the id of the cluster, to which the gateway
573 --target_cluster_id\SpecialChar ~
574 id Specify the id of the cluster, to which the target
575 of the route belongs.
579 nid For a range route, this is the lo value
585 nid For a range route, this is the hi value
592 echo-client Used for testing purpose only.
602 Node on which the echo-client resides.
608 \labelwidthstring 00.00.0000
615 --reference Print short reference for commands.
618 --verbose Print system commands as they are run.
621 --batch\SpecialChar ~
622 filename Used to execute lmc commands in batch mode.
625 --output\SpecialChar ~
626 filename Send output to the file.
627 If the file exists, it will be overwritten.
630 --merge\SpecialChar ~
635 Add the new element to an existing file.
640 \layout Subsubsection
647 lmc --node adev3 --add net --nid adev3 --cluster_id 0x1000 --nettype tcp
648 --hostaddr adev3-eth0 --port 988
653 lmc --node adev3 --add net --nid adev3 --cluster_id 0x2000 --nettype tcp
654 --hostaddr adev3-eth1 --port 989
660 Used to add a Lustre node to a specified Lustre cluster through a network
662 In this example, Lustre node adev3 has been added to 2 Lustre clusters
663 whose cluster_id are 0x1000 and 0x2000 separately through 2 network interface
664 cards: adev3-eth0 and adev3-eth1.
665 adev3 would listen in some specified port(s) to prepare for possible connection
666 requests from nodes in these two clusters.
667 \layout Subsubsection
674 lmc --node adev3 --add route --nettype tcp --gw 5 --gateway_cluster_id 0x1000
675 --target_cluster_id 0x1000 --lo 4 --hi 7
678 Used to add a route entry for a Lustre node.
679 Here Lustre node adev3 is told it has a new route entry that now it could
680 send packets to Lustre nodes whose nids is from 4 to 7 with the help of
681 Lustre gateway node whose nid is 5.
682 Besides, Lustre gateway node is in cluster whose id is 0x1000 and target
683 of the route belongs to cluster whose id is also 0x1000.
684 The network in this route path is a tcp network.