vzctl restore: Error: undump failed: File exists [message #42615] |
Fri, 06 May 2011 14:34 |
divB
Messages: 79 Registered: April 2009
|
Member |
|
|
Hi,
When doing online migration, I get the following error: Error: undump failed: File exists. Full output:
# vzmigrate -r no --keep-dst --online -v 10.7.222.2 301
Starting online migration of CT 301 to 10.7.222.2
OpenVZ is running...
Loading /etc/vz/vz.conf and /etc/vz/conf/301.conf files
Check IPs on destination node:
Preparing remote node
Copying config file
301.conf 100% 1730 1.7KB/s 00:00
Name mail assigned
Saved parameters for CT 301
Creating remote container root dir
Creating remote container private dir
VZ disk quota disabled -- skipping quota migration
Syncing private
Live migrating container...
Suspending container
Setting up checkpoint...
suspend...
get context...
Checkpointing completed succesfully
Dumping container
Setting up checkpoint...
join context..
dump...
Checkpointing completed succesfully
Copying dumpfile
dump.301 100% 187MB 37.5MB/s 00:05
Syncing private (2nd pass)
VZ disk quota disabled -- skipping quota migration
Undumping container
Restoring container ...
Starting container ...
Container is mounted
undump...
Setting CPU units: 11222
Configure veth devices: veth301.0
Adding interface veth301.0 to bridge br-wan on CT0 for CT301
Error: undump failed: File exists
Restoring failed:
Error: open_deleted: no suffix
Error: rst_file: -17 308640
Error: rst_files: -17
Error: make_baby: -17
Error: rst_clone_children
Error: make_baby: -17
Error: rst_clone_children
Stopping container ...
Container start failed
Container was stopped
Container is unmounted
Error: Failed to undump container
Resuming...
This error is output from "vzctl restore" command on this line of vzmigrate script:
logexec 2 $SSH root@$host vzctl restore $VEID --undump --dumpfile $VE_DUMPFILE --skip_arpdetect
Any hints/help is greatly appreciated!
Regards,
divB
/EDIT: I tried around with small and big containers ... very often I get some errors: File exists, no space left on device, ... Sometimes it helps to reboot the destination hardware node, sometimes it helps to restart specific services within a container, sometimes certain processes need to be killed before migration. In fact, it is pure random whether live migration works. It seems to be very error prone and mostly unuseable in practice. I am the one person who has these problems? There is very few information on google
[Updated on: Fri, 06 May 2011 16:48] Report message to a moderator
|
|
|
|
|