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, Jun 20, 2017 at 01:47:08PM +0000, Dominic Pratt wrote: > Hi there, > > we're running DRBD9/PVE4, created two disks on DRBD yesterday and there were no errors at all. > > Trying to start that VM up, we recognized, that there was no link to /dev/drbd/by-res/. The result is, that the VM cannot start up at all. The link is only missing on one of the cluster nodes, on the other two the link has been created. > > Checking the syslog, we recognized these lines: > > Jun 20 08:41:00 px1 systemd-udevd[11667]: timeout '/usr/sbin/drbdadm sh-udev minor-131' > Jun 20 08:41:00 px1 systemd-udevd[11667]: timeout '/lib/udev/vdev_id -d drbd131' > Jun 20 08:41:02 px1 systemd-udevd[11663]: timeout '/usr/sbin/drbdadm sh-udev minor-130' > Jun 20 08:41:02 px1 systemd-udevd[11663]: timeout '/lib/udev/vdev_id -d drbd130' > > It seems, that udev timed out and drbd does not check it's exit status? This happened even though we had the "sleep 10" in our drbd-code as stated in our mails on here before. Uhm, it's the other way around. *udev* is executing drbdadm, and *udev* (specifically, systemd udev here) thinks that executing that helper program took too long. -- : Lars Ellenberg : LINBIT | Keeping the Digital World Running : DRBD -- Heartbeat -- Corosync -- Pacemaker DRBD® and LINBIT® are registered trademarks of LINBIT __ please don't Cc me, but send to list -- I'm subscribed