Hello, i am trying to use lustre as per examples but i am getting following error. Oct 24 15:13:49 news6 kernel: LustreError: 9799:0:(fsfilt-ldiskfs.c:339:fsfilt_ldiskfs_brw_start()) Skipped 152 previous similar messages Oct 24 15:13:49 news6 kernel: LustreError: 9799:0:(filter_io_26.c:646:filter_commitrw_write()) error starting transaction: rc = -30 Oct 24 1 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.clusterfs.com/pipermail/lustre-discuss/attachments/20061024/208aef86/attachment.html
The -30 is the Error EROFS or Read-only file system. Are you getting this error on an OSS, or on your MDS? Either way look back further in your logs for some other reason the file system went read-only. The default action if ext3/ldiskfs detects any type of error or corruption on the file system is to re-mount read-only, to protect against further corruption. Evan ________________________________ From: lustre-discuss-bounces@clusterfs.com [mailto:lustre-discuss-bounces@clusterfs.com] On Behalf Of Kuldeep Kr Sent: Tuesday, October 24, 2006 6:24 AM To: Lustre-discuss@clusterfs.com Subject: [Lustre-discuss] Lustre Over Storage Hello, i am trying to use lustre as per examples but i am getting following error. Oct 24 15:13:49 news6 kernel: LustreError: 9799:0:(fsfilt-ldiskfs.c:339:fsfilt_ldiskfs_brw_start()) Skipped 152 previous similar messages Oct 24 15:13:49 news6 kernel: LustreError: 9799:0:(filter_io_26.c:646:filter_commitrw_write()) error starting transaction: rc = -30 Oct 24 1 <http://adworks.rediff.com/cgi-bin/AdWorks/sigclick.cgi/www.rediff.com/s ignature-home.htm/1507191490@Middle5?PARTNER=3> -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.clusterfs.com/pipermail/lustre-discuss/attachments/20061024/28d931cd/attachment.html
? Actually what is happening .. I am storing data regularly on my lustre partition but when the size of lustre partition becomes equal to my free physical storgae then it gives this error. my lustre patition is showing that i have a lot free space but my physical storage is full as the size of OST is increasing rapidly as i save data on my lustre partition and when i am deleting data from lustre it make /lustre free but the size of OST is still same. Can u guide how can i do this ???? Thanks On Tue, 24 Oct 2006 Felix,Evan J wrote :>The -30 is the Error EROFS or Read-only file system. Are you getting >this error on an OSS, or on your MDS? Either way look back further in >your logs for some other reason the file system went read-only. The >default action if ext3/ldiskfs detects any type of error or corruption >on the file system is to re-mount read-only, to protect against further >corruption. > >Evan > > >________________________________ > > From: lustre-discuss-bounces@clusterfs.com >[mailto:lustre-discuss-bounces@clusterfs.com] On Behalf Of Kuldeep Kr > Sent: Tuesday, October 24, 2006 6:24 AM > To: Lustre-discuss@clusterfs.com > Subject: [Lustre-discuss] Lustre Over Storage > > > > Hello, > i am trying to use lustre as per examples but i am getting >following error. > Oct 24 15:13:49 news6 kernel: LustreError: >9799:0:(fsfilt-ldiskfs.c:339:fsfilt_ldiskfs_brw_start()) Skipped 152 >previous similar messages > Oct 24 15:13:49 news6 kernel: LustreError: >9799:0:(filter_io_26.c:646:filter_commitrw_write()) error starting >transaction: rc = -30 > Oct 24 1 > > > > ><http://adworks.rediff.com/cgi-bin/AdWorks/sigclick.cgi/www.rediff.com/s >ignature-home.htm/1507191490@Middle5?PARTNER=3> >+39-328-175-4828 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.clusterfs.com/pipermail/lustre-discuss/attachments/20061024/53f760e8/attachment.html
On Oct 24, 2006 15:27 -0000, Kuldeep Kr wrote:> ? > Actually what is happening .. I am storing data regularly on my lustre partition but when the size of lustre partition becomes equal to my free physical storgae then it gives this error. my lustre patition is showing that i have a lot free space but my physical storage is full as the size of OST is increasing rapidly as i save data on my lustre partition and when i am deleting data from lustre it make /lustre free but the size of OST is still same. Can u guide how can i do this ????As I previously replied to your other emails - this is caused by the use of loopback devices instead of real block devices for your MDS and OST. I suspect you have created sparse loopback files that are larger than your actual filesystem, and once the backing filesystem is full it will cause problems even if there is free space inside the loopback filesystem. Cheers, Andreas -- Andreas Dilger Principal Software Engineer Cluster File Systems, Inc.
Ok.. so loopback files should be smaller or eqal to actual storage ? now my loopback is smaller then actual storage but still have following error.. any suggestion please ? Oct 26 09:31:43 news7 kernel: LustreError: 8963:0:(client.c:940:ptlrpc_expire_one_request()) Skipped 8 previous similar messages Oct 26 09:31:43 news7 kernel: LustreError: OSC_news7.telebyte.nl_OST_news7.telebyte.nl_MNT_news7.telebyte.nl: Connection to service OST_news7.telebyte.nl via nid 0@lo was lost; in progress operations using this service will wait for recovery to complete. Oct 26 09:31:43 news7 kernel: LustreError: 8963:0:(client.c:940:ptlrpc_expire_one_request()) @@@ timeout (sent at 1161847802, 100s ago) req@cdf1f600 x11561/t0 o6->OST_news7.telebyte.nl_UUID@news7.telebyte.nl_UUID:28 lens 280/280 ref 1 fl Rpc:/0/0 rc 0/0 On Tue, 24 Oct 2006 Andreas Dilger wrote :>On Oct 24, 2006 15:27 -0000, Kuldeep Kr wrote: > > > > Actually what is happening .. I am storing data regularly on my lustre partition but when the size of lustre partition becomes equal to my free physical storgae then it gives this error. my lustre patition is showing that i have a lot free space but my physical storage is full as the size of OST is increasing rapidly as i save data on my lustre partition and when i am deleting data from lustre it make /lustre free but the size of OST is still same. Can u guide how can i do this ???? > >As I previously replied to your other emails - this is caused by the use >of loopback devices instead of real block devices for your MDS and OST. > >I suspect you have created sparse loopback files that are larger than >your actual filesystem, and once the backing filesystem is full it will >cause problems even if there is free space inside the loopback filesystem. > >Cheers, Andreas >-- >Andreas Dilger >Principal Software Engineer >Cluster File Systems, Inc. >+39-328-175-4828 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.clusterfs.com/pipermail/lustre-discuss/attachments/20061026/6ac8f45c/attachment-0001.html