One more question, what's the behavior behavior if the disk which contains every journal crashed? All the OSDs will crash? Thank you in advance :) On Mon, Aug 27, 2012 at 7:37 PM, Sébastien Han <han.sebastien@xxxxxxxxx> wrote: > Ok, I see. Thanks for the clarification :) > > > On Mon, Aug 27, 2012 at 6:20 PM, Gregory Farnum <greg@xxxxxxxxxxx> wrote: >> On Mon, Aug 27, 2012 at 9:10 AM, Sébastien Han <han.sebastien@xxxxxxxxx> wrote: >>> Ok for the lazy umount but the journal is not here anymore, maybe >>> still written in memory. >> >> The lazy unmount keeps the filesystem attached to the tree until >> nobody's using it anymore. What you've done here is to say "as soon as >> nobody's touching this, kill it!" So the Ceph OSD keeps on using the >> journal, and Linux keeps it around...and then when you shut down the >> OSD, Linux removes the journal (*after* the OSD is done with it). >> There is literally nothing that Ceph can do about this scenario. >> -Greg >> >>> On Mon, Aug 27, 2012 at 10:59 AM, Jerker Nyberg <jerker@xxxxxxxxxxxx> wrote: >>>> On Sun, 26 Aug 2012, Sébastien Han wrote: >>>> >>>>> I used: >>>>> >>>>> umount -l to bypassed the warning. >>>> >>>> >>>> "umount -l" does a lazy unmount. The file system is not really unmounted >>>> until it is not busy anymore. Ceph may continue to use the open journals and >>>> they should be visible in the file system hierarchy. >>>> >>>> Kind regards, >>>> Jerker Nyberg. >>> -- >>> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in >>> the body of a message to majordomo@xxxxxxxxxxxxxxx >>> More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
