At my university''s department we have a policy to reboot our server on a bi-weekly basis. I was wondering if anyone does anything special in their Lustre environment. We currently just run a simple reboot on the storage hosts (OSTS and MDS), and then the clients. My question is, is there anything special I must do on the client side? or the storage side so nothing freezes and my reboot goes smooth? TIA
Anyone have any thoughts on this issue? My clients are still freezing for some reason... On Sat, Aug 23, 2008 at 8:54 AM, Mag Gam <magawake at gmail.com> wrote:> At my university''s department we have a policy to reboot our server on > a bi-weekly basis. I was wondering if anyone does anything special in > their Lustre environment. We currently just run a simple reboot on the > storage hosts (OSTS and MDS), and then the clients. My question is, is > there anything special I must do on the client side? or the storage > side so nothing freezes and my reboot goes smooth? > > TIA >
Just a guess - are you tearing down the network before unmounting the lustre mount on the client? Mag Gam wrote:> Anyone have any thoughts on this issue? > > My clients are still freezing for some reason... > > > On Sat, Aug 23, 2008 at 8:54 AM, Mag Gam <magawake at gmail.com> wrote: > >> At my university''s department we have a policy to reboot our server on >> a bi-weekly basis. I was wondering if anyone does anything special in >> their Lustre environment. We currently just run a simple reboot on the >> storage hosts (OSTS and MDS), and then the clients. My question is, is >> there anything special I must do on the client side? or the storage >> side so nothing freezes and my reboot goes smooth? >> >> TIA >> >> > _______________________________________________ > Lustre-discuss mailing list > Lustre-discuss at lists.lustre.org > http://lists.lustre.org/mailman/listinfo/lustre-discuss > >
I think I am. But if I try to unmount it near impossible because I still have users accessing the lustre filesystem. I was wondering if there was a lctl command I can run from the MDS to simply stop all clients. On Mon, Aug 25, 2008 at 12:01 AM, Adam Cassar <adam.cassar at netregistry.com.au> wrote:> Just a guess - are you tearing down the network before unmounting the > lustre mount on the client? > > Mag Gam wrote: >> Anyone have any thoughts on this issue? >> >> My clients are still freezing for some reason... >> >> >> On Sat, Aug 23, 2008 at 8:54 AM, Mag Gam <magawake at gmail.com> wrote: >> >>> At my university''s department we have a policy to reboot our server on >>> a bi-weekly basis. I was wondering if anyone does anything special in >>> their Lustre environment. We currently just run a simple reboot on the >>> storage hosts (OSTS and MDS), and then the clients. My question is, is >>> there anything special I must do on the client side? or the storage >>> side so nothing freezes and my reboot goes smooth? >>> >>> TIA >>> >>> >> _______________________________________________ >> Lustre-discuss mailing list >> Lustre-discuss at lists.lustre.org >> http://lists.lustre.org/mailman/listinfo/lustre-discuss >> >> > _______________________________________________ > Lustre-discuss mailing list > Lustre-discuss at lists.lustre.org > http://lists.lustre.org/mailman/listinfo/lustre-discuss >