We previously had a filesystem and we removed its MGS/MDS and its OSTS. We created a new filesystem and everything is fine now. However, We still see messages like this: "Lustre Error 137-5: UUID ''oldfs001-OST0000-UUID'' is not available for connect (no target) '' On the MDS, i even did a ''lctl dl | grep oldfs'' and no entries come up. Is there something I am missing? TIA
Using lustre 1.6.4.3 with kernel 2.6.22 TIA On Thu, Mar 19, 2009 at 10:55 PM, Mag Gam <magawake at gmail.com> wrote:> We previously had a filesystem and we removed its MGS/MDS and its > OSTS. We created a new filesystem and everything is fine now. However, > We still see messages like this: > "Lustre Error 137-5: UUID ''oldfs001-OST0000-UUID'' is not available for > connect (no target) '' > > > On the MDS, i even did a ''lctl dl | grep oldfs'' and no entries come up. > > Is there something I am missing? > > TIA >
On Thu, 2009-03-19 at 22:55 -0400, Mag Gam wrote:> > "Lustre Error 137-5: UUID ''oldfs001-OST0000-UUID'' is not available for > connect (no target) ''This message simply means that a lustre client has come to this node looking for a target and it''s not active on this server. Not sure what it means for your environment given your recent activity (maybe just a client with the old filesystem still mounted), but typically in an otherwise stable environment it can indicate networking errors as a client has had opportunity to try to find a service on the non-active server, meaning that it''s connection to a service was interrupted. This is of course normal when a client first mounts a filesystem as it tries to find the active server. b, -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 197 bytes Desc: This is a digitally signed message part Url : http://lists.lustre.org/pipermail/lustre-discuss/attachments/20090320/94ed9273/attachment-0001.bin
in other words, reboot all of your clients? On Fri, Mar 20, 2009 at 12:25 PM, Brian J. Murrell <Brian.Murrell at sun.com> wrote:> On Thu, 2009-03-19 at 22:55 -0400, Mag Gam wrote: >> >> "Lustre Error 137-5: UUID ''oldfs001-OST0000-UUID'' is not available for >> connect (no target) '' > > This message simply means that a lustre client has come to this node > looking for a target and it''s not active on this server. > > Not sure what it means for your environment given your recent activity > (maybe just a client with the old filesystem still mounted), but > typically in an otherwise stable environment it can indicate networking > errors as a client has had opportunity to try to find a service on the > non-active server, meaning that it''s connection to a service was > interrupted. > > This is of course normal when a client first mounts a filesystem as it > tries to find the active server. > > b, > > > _______________________________________________ > Lustre-discuss mailing list > Lustre-discuss at lists.lustre.org > http://lists.lustre.org/mailman/listinfo/lustre-discuss > >
On Sat, 2009-03-21 at 02:22 -0400, Mag Gam wrote:> in other words, reboot all of your clients?No. Not so much. If clients have been set to mount that removed filesystem on reboot, then you will be right back where you started. Although, in fact, they will probably get an error trying to mount (because it doesn''t exist any more) but you will still have clients that you need to fix in that case. b. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 197 bytes Desc: This is a digitally signed message part Url : http://lists.lustre.org/pipermail/lustre-discuss/attachments/20090323/c0821d0c/attachment.bin
Thanks Brian. This answers a lot of questions. Thanks for the great insight. On Mon, Mar 23, 2009 at 8:45 AM, Brian J. Murrell <Brian.Murrell at sun.com> wrote:> On Sat, 2009-03-21 at 02:22 -0400, Mag Gam wrote: >> in other words, reboot all of your clients? > > No. Not so much. If clients have been set to mount that removed > filesystem on reboot, then you will be right back where you started. > Although, in fact, they will probably get an error trying to mount > (because it doesn''t exist any more) but you will still have clients that > you need to fix in that case. > > b. > > > _______________________________________________ > Lustre-discuss mailing list > Lustre-discuss at lists.lustre.org > http://lists.lustre.org/mailman/listinfo/lustre-discuss > >