Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On Tue, Oct 06, 2009 at 02:31:40PM +0200, Rainer Traut wrote: > Lars Ellenberg schrieb: >> On Tue, Oct 06, 2009 at 12:53:20PM +0200, Rainer Traut wrote: >>> I see this in dmesg with 8.3.3 and RHEL5.4 x86_64: >>> >>> block drbd1: write: error=-95 s=5767576s >>> block drbd1: Method to ensure write ordering: flush >>> >>> while in earlier drbd version 8.3.2: >>> >>> kernel: block drbd1: Method to ensure write ordering: barrier >>> >>> Anything to be concerned? >> >> Nope. >> 95 is EOPNOTSUPP, barriers seem to be not supported on your lower level >> device, so DRBD falls back to "flush" mode. >> which, because implemented as "empty barriers" probably fail as well, >> so I guess it falls back further to "drain" mode pretty quickly. >> >> That should be no different between 8.3.2 and 8.3.3, though. >> > > Yeah, nevertheless this has changed between 8.3.2 and 8.3.3 - only > change was drbd update. > > Sys: kernel 2.6.18-164.2.1.el5 x86_64 > > drbd on top of LVM: > device /dev/drbd1; > disk /dev/VG00/LVvm1slash; I'm _very_ certain that, if you go back to 8.3.2, you'd see the same behaviour, and most likely it has been that way before, as well, or the upgrade was _not_ the only change, but you changed some kernel as well, or you changed the layout of your LV (lvextent? pvmove?). Anyways: no worries. -- : Lars Ellenberg : LINBIT HA-Solutions GmbH : DRBD®/HA support and consulting http://www.linbit.com DRBD® and LINBIT® are registered trademarks of LINBIT, Austria. __ please don't Cc me, but send to list -- I'm subscribed