DRBD: In transient state log entry when live migrating
32 views
Skip to first unread message
John N.
Sep 4, 2018, 1:05:02 PM9/4/18
to ganeti
Hello,
From time to time I check my ganeti node-daemon.log log file especially when doing live migrations from one node to the other. Now I noticed that when live migrating some ganeti instances I get the following error and warning:
2018-09-04 17:01:13,657: ganeti-noded pid=7618 INFO RunCmd drbdsetup disconnect ipv4:10.0.2.12:11026 ipv4:10.0.2.13:11026
2018-09-04 17:01:13,672: ganeti-noded pid=7618 ERROR drbd26: can't shutdown network: ??: State change failed: (-18) In transient state, retry after next state change
2018-09-04 17:01:13,672: ganeti-noded pid=7618 WARNING Caught BlockDeviceError but ignoring: drbd26: can't shutdown network: ??: State change failed: (-18) In transient state, retry after next state change
2018-09-04 17:01:13,673: ganeti-noded pid=7618 INFO RunCmd drbdsetup disconnect ipv4:10.0.2.12:11026 ipv4:10.0.2.13:11026
The "gnt-instance migrate" does not report any errors and the DRBD disks on both nodes (primary and secondary) seem fine (by checking /proc/drbd). So I was wondering if anyone of you also noticed this error message and if anyone knows if this is bad or if it can safely be ignored...
My setup is a 3 nodes cluster with ganeti 2.12 official debian packages on Debian 8 running Xen 4.4 as hypervisor with DRBD redundant instances.
Best regards,
John
From time to time I check my ganeti node-daemon.log log file especially when doing live migrations from one node to the other. Now I noticed that when live migrating some ganeti instances I get the following error and warning:
2018-09-04 17:01:13,657: ganeti-noded pid=7618 INFO RunCmd drbdsetup disconnect ipv4:10.0.2.12:11026 ipv4:10.0.2.13:11026
2018-09-04 17:01:13,672: ganeti-noded pid=7618 ERROR drbd26: can't shutdown network: ??: State change failed: (-18) In transient state, retry after next state change
2018-09-04 17:01:13,672: ganeti-noded pid=7618 WARNING Caught BlockDeviceError but ignoring: drbd26: can't shutdown network: ??: State change failed: (-18) In transient state, retry after next state change
2018-09-04 17:01:13,673: ganeti-noded pid=7618 INFO RunCmd drbdsetup disconnect ipv4:10.0.2.12:11026 ipv4:10.0.2.13:11026
The "gnt-instance migrate" does not report any errors and the DRBD disks on both nodes (primary and secondary) seem fine (by checking /proc/drbd). So I was wondering if anyone of you also noticed this error message and if anyone knows if this is bad or if it can safely be ignored...
My setup is a 3 nodes cluster with ganeti 2.12 official debian packages on Debian 8 running Xen 4.4 as hypervisor with DRBD redundant instances.
Best regards,
John
Reply all
Reply to author
Forward
0 new messages