Talk:Lustre (file system)

GPFS vs Lustre
Lustre does not directly talk or manage OSTs, and delegates this responsibility to OSSs. But the OSSs are a part of Lustre, aren't they? (as well as the OSTs). Do you mean The MDS does not directly talk to OSTs? --140.181.85.77 14:05, 8 June 2007 (UTC)

A: The distinction is that in a shared block filesystem like GPFS all of the clients write directly to the disks in the blocks they are assigned, while in Lustre the access to the disks is by (essentially) {object, offset, size} and is moderated by the OSS RPC interface. In the shared-block case, any kind of programming error (or malicious behaviour) on behalf of the client can read or overwrite arbitrary parts of the disk, while in Lustre the OSS can authenticate the client, the user, and the operations it is performing on each file in the filesystem. 70.72.213.136 09:07, 18 October 2007 (UTC)

I changed the page so that is identifies Lustre as a distributed file system (like Ceph, GlusterFS, NFS etc) not a shared disk file system (like OCFS, GFS etc). As far as I undestand the clients in Lustre do not have block level access to the block storage on which the file system is located. --JerkerNyberg (talk) 10:08, 9 November 2008 (UTC)

Fair use rationale for Image:Lustre logo.gif
Image:Lustre logo.gif is being used on this article. I notice the image page specifies that the image is being used under fair use but there is no explanation or rationale as to why its use in this Wikipedia article constitutes fair use. In addition to the boilerplate fair use template, you must also write out on the image description page a specific explanation or rationale for why using this image in each article is consistent with fair use.

Please go to the image description page and edit it to include a fair use rationale. Using one of the templates at Fair use rationale guideline is an easy way to insure that your image is in compliance with Wikipedia policy, but remember that you must complete the template. Do not simply insert a blank template on an image page.

If there is other fair use media, consider checking that you have specified the fair use rationale on the other images used on this page. Note that any fair use images lacking such an explanation can be deleted one week after being tagged, as described on criteria for speedy deletion. If you have any questions please ask them at the Media copyright questions page. Thank you.

BetacommandBot (talk) 21:36, 2 January 2008 (UTC)

DataDirect Reference
The DataDirect link is invalid. I'd pull the reference but as it is needed to back the sentence, we really need a replacement. Does anyone have any links to other articles on data-directed computation on Lustre? —Preceding unsigned comment added by SteveLoughran (talk • contribs) 11:18, 7 October 2008 (UTC)

Filesystem Infobox
This page uses the Software infobox. However, it seems to be equally appropriate to use the Filesystem Infobox, which provides different fields of information. Is it possible to use both Infoboxes? Could some fields be added to Filesystem Infobox such that the Software Infobox would not be necessary?. — Preceding unsigned comment added by 207.219.56.130 (talk) 18:48, 19 May 2015 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified 4 external links on Lustre (file system). Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20141031152502/http://zfsonlinux.org/docs/LUG11_ZFS_on_Linux_for_Lustre.pdf to http://zfsonlinux.org/docs/LUG11_ZFS_on_Linux_for_Lustre.pdf
 * Added archive https://archive.is/20130203232617/http://www.taborcommunications.com/dsstar/03/1125/107031.html to http://www.taborcommunications.com/dsstar/03/1125/107031.html
 * Added archive https://www.webcitation.org/65tGGL6n1?url=http://hepix.caspur.it/afs/hepix.org/project/strack/hep_pdf/2007/Spring/Lustre-CEA-hepix2007.pdf to http://hepix.caspur.it/afs/hepix.org/project/strack/hep_pdf/2007/Spring/Lustre-CEA-hepix2007.pdf
 * Added archive https://web.archive.org/web/20160621213259/http://www.xyratex.com/products/lustre to https://www.xyratex.com/products/lustre

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 07:20, 9 January 2018 (UTC)

Network throughput claims
LNet provides end-to-end throughput over Gigabit Ethernet networks in excess of 100 MB/s,[75] throughput up to 11 GB/s using InfiniBand enhanced data rate (EDR) links, and throughput over 11 GB/s across 100 Gigabit Ethernet interfaces. The current phrasing about throughput seems fairly irrelevant, Lustre (and other clustered filesystems) can saturate the links of multiple, be it gigabit Ethernet or 200G EDR IB, clients provided the setup of hardware and software accommodates it, in a small setup or with slow disks on the other hand the limiting factor will be said disks. With a decent single NFS server one may even be able to saturate its link under the right circumstances, of course with multiple clients this situation would change. --Keeper of the Keys (talk) 22:17, 25 June 2020 (UTC)