Hi, Cliff White. Oh, yes. I'm mounting the client with -o flock. The problem with it right? thanks,
On Tue, Oct 6, 2009 at 1:25 AM, Cliff White <[email protected]> wrote: > Đào Thị Thảo wrote: > >> Hello all, >> I have a problem with lustre 1.8.0 on Centos 5.0. >> >> There are logs on client: >> Oct 4 23:56:13 web5 kernel: Lustre: lustre-MDT0000-mdc-ffff81022e8e9400: >> Connection to service >> > <snip> > >> >> It's look like the disconnect tcp between MDS and client, but i check lctl >> ping, it's OK. The except, MDS (or other nodes) that can't ssh to this >> client. My provisional measure is reboot this client. It works OK. But to a >> certain time, it happens again ( can happen with other clients). >> I don't think, the cause is network. Because, my network is still stable. >> what's happen with my system and how to fix it? >> I know the problem client evicted has been discussed a lot, but I do not >> find a way to solve my problem. >> thanks, >> >> > Are you mounting the client with -o flock ? > > cliffw > > >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Lustre-discuss mailing list >> [email protected] >> http://lists.lustre.org/mailman/listinfo/lustre-discuss >> > >
_______________________________________________ Lustre-discuss mailing list [email protected] http://lists.lustre.org/mailman/listinfo/lustre-discuss
