Talk:Lustre FS: Difference between revisions
(New page: What happens if one of the sharing members "drops out" unexpectedly? This problem is really hard. I would think it would probably be easier to just assign group scratch space on the server...) |
No edit summary |
||
| Line 2: | Line 2: | ||
[[User:Homunq|Homunq]] 13:40, 4 April 2008 (EDT) |
[[User:Homunq|Homunq]] 13:40, 4 April 2008 (EDT) |
||
In case of today's LustreFS implementation it will lead to loss of data. However, having added mirroring, it is possible to avoid the this problem. Your decision is undoubtedly quickly sold and is in any measure reliable. Port of LustreFS will allow to make a step in a direction of access and storage of a content of great volume on devices with the limited disk space. As hardware is fixed, the problem of storage of greater data have to be solved in program way, for example, software RAID. Port of LustreFS is the first attempt to make it . |
|||
[[User:Elka| Evelina Stepanova]] 23:39, 7 April 2008 (MSK) |
|||
Latest revision as of 19:39, 7 April 2008
What happens if one of the sharing members "drops out" unexpectedly? This problem is really hard. I would think it would probably be easier to just assign group scratch space on the server drive, and mount that over IP. Every large XO installation will have a school server, even if it doesn't have internet.
Homunq 13:40, 4 April 2008 (EDT)
In case of today's LustreFS implementation it will lead to loss of data. However, having added mirroring, it is possible to avoid the this problem. Your decision is undoubtedly quickly sold and is in any measure reliable. Port of LustreFS will allow to make a step in a direction of access and storage of a content of great volume on devices with the limited disk space. As hardware is fixed, the problem of storage of greater data have to be solved in program way, for example, software RAID. Port of LustreFS is the first attempt to make it .
Evelina Stepanova 23:39, 7 April 2008 (MSK)