Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On 05/19/2016 08:58 AM, d tbsky wrote: > 2016-05-19 23:35 GMT+08:00 Lars Ellenberg <lars.ellenberg at linbit.com>: >> Even though the subject does not mention drbd online verify, >> you still want to read this: >> "Re: Explained: Digest integrity check FAILED" >> http://lists.linbit.com/pipermail/drbd-user/2011-March/015751.html >> http://www.gossamer-threads.com/lists/drbd/users/21091?do=post_view_threaded#21091 >> or gmane or any other ML archive. >> >> you could run with data-integrity enabled for a while, and check for >> "Digest mismatch, buffer modified by upper layers during write" >> messages in kernel logs. > > hi: > I already try to enable data-integrity check after the > verify/resync failed, but it didn't help. the verify/resync was not > correct and there are no "Digest mismatch" messages in kernel logs. > > I now will write a script to verfiy/resync the resource volumes > one by one, to see if all the resources volumes can be verify/synced > correctly. > > or if you have other ideas need to try at the cluster please tell > me. the cluster is running many testing vms, so the corruption is > affordable for these vms. It could very well be a sort of race condition like you mentioned earlier. If the block is written on the Primary, but the verify is checksumming the backing disks before the write makes it down to disk on the Secondary (block is on the wire, in a buffer, etc), the verify could be coming up with false positives. This is described in one of LINBIT's blog posts found here: https://blogs.linbit.com/p/138/trust-but-verify/ > > thanks a lot for your help!! > > Regards, > tbskyd > _______________________________________________ > drbd-user mailing list > drbd-user at lists.linbit.com > http://lists.linbit.com/mailman/listinfo/drbd-user > -- : Matt Kereczman : matt at linbit.com : +1 (503) 573-1262 x205 : LINBIT | Keeping the Digital World Running : DRBD -- Heartbeat -- Corosync -- Pacemaker