Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On Thu, Apr 26, 2007 at 10:32:53AM -0400, Charles Bennington wrote: > Just running fsck on /dev/drbd0 will require bringing the cluster off > lines so that I can umount the filesystem -- which isn't really an > option. So it sounds like I really am stuck with taking one node off > line to perform the fsck and then switching over to it so that it > becomes master and then re-syncing the other node? do you need to write to this filesystem ? what happens to your writes in the above scenario ? perhaps someone should roll up their sleeves and repair the filesystem while the system is live ? Regards, Paddy