1Lustre Parallel Filesystem Client 2================================= 3 4The Lustre file system is an open-source, parallel file system 5that supports many requirements of leadership class HPC simulation 6environments. 7Born from from a research project at Carnegie Mellon University, 8the Lustre file system is a widely-used option in HPC. 9The Lustre file system provides a POSIX compliant file system interface, 10can scale to thousands of clients, petabytes of storage and 11hundreds of gigabytes per second of I/O bandwidth. 12 13Unlike shared disk storage cluster filesystems (e.g. OCFS2, GFS, GPFS), 14Lustre has independent Metadata and Data servers that clients can access 15in parallel to maximize performance. 16 17In order to use Lustre client you will need to download lustre client 18tools from 19https://downloads.hpdd.intel.com/public/lustre/latest-feature-release/ 20the package name is lustre-client. 21 22You will need to install and configure your Lustre servers separately. 23 24Mount Syntax 25============ 26After you installed the lustre-client tools including mount.lustre binary 27you can mount your Lustre filesystem with: 28 29mount -t lustre mgs:/fsname mnt 30 31where mgs is the host name or ip address of your Lustre MGS(management service) 32fsname is the name of the filesystem you would like to mount. 33 34 35Mount Options 36============= 37 38 noflock 39 Disable posix file locking (Applications trying to use 40 the functionality will get ENOSYS) 41 42 localflock 43 Enable local flock support, using only client-local flock 44 (faster, for applications that require flock but do not run 45 on multiple nodes). 46 47 flock 48 Enable cluster-global posix file locking coherent across all 49 client nodes. 50 51 user_xattr, nouser_xattr 52 Support "user." extended attributes (or not) 53 54 user_fid2path, nouser_fid2path 55 Enable FID to path translation by regular users (or not) 56 57 checksum, nochecksum 58 Verify data consistency on the wire and in memory as it passes 59 between the layers (or not). 60 61 lruresize, nolruresize 62 Allow lock LRU to be controlled by memory pressure on the server 63 (or only 100 (default, controlled by lru_size proc parameter) locks 64 per CPU per server on this client). 65 66 lazystatfs, nolazystatfs 67 Do not block in statfs() if some of the servers are down. 68 69 32bitapi 70 Shrink inode numbers to fit into 32 bits. This is necessary 71 if you plan to reexport Lustre filesystem from this client via 72 NFSv4. 73 74 verbose, noverbose 75 Enable mount/umount console messages (or not) 76 77More Information 78================ 79You can get more information at 80OpenSFS website: http://lustre.opensfs.org/about/ 81Intel HPDD wiki: https://wiki.hpdd.intel.com 82 83Out of tree Lustre client and server code is available at: 84http://git.whamcloud.com/fs/lustre-release.git 85 86Latest binary packages: 87http://lustre.opensfs.org/download-lustre/ 88