Restore VM after hypervisor crash / WRITE EPIPE - IMPORT_INCOMPATIBLE_VERSION()

anivard

Hello,

After a crash server, only disk system.
I reinstalled XCP-NG.
To restore my VM, I have to install XO, done.
But after that, I can't restore a VM through XO.

I need to restore my Firewall to have the correct setting of my network but I have WRITE EPIPE. My server and my backup NAS are on the same subnet with the correct access right.

I would like also restore my old XOA because of settings are on it but IMPORT_INCOMPATIBLE_VERSION()
PS: I backup the XOA configuration via the backup process in XO.

I attached the log because I don't understand why after a complete reinstall of an hypervisor, it's difficult to restart from backup.
firewall.txt
xoa.txt

anivard

Hello again,

I don't know what's happend but now I can restore my VM!

Before this success, I reinstalled with 2 VM and I created a DR job to replicate my DR VM on my NAS. Perhaps, it's unlock something.

Now, I try to mount my ZFS volume but no success...

SR_BACKEND_FAILURE_61(, File SR creation error, )

I re-imported the zpool with this command :

zpool import -a

I can this status pool is good and I can this all VHD files.

This is my new "fight" !

anivard

I try 2 things :

1/ Create SR when the zfs pool is mount

zpool import -a

Via the XO NEW + STORAGE > ZFS

"SR_BACKEND_FAILURE_61(, File SR creation error, )"

2/ Create SR when the zfs pool is unmount
After the

zpool import -a

umount /mntzpool

Via command line on XCP-ng server

xe sr-create host-uuid=4dab3683-0cbf-4b01-a2fb-49989171c596 type=zfs content-type=user name-label=Local4TBraidz device-config:location=/local4TBraidz/

Error code: SR_BACKEND_FAILURE_202
Error parameters: , General backend error [opterr=Command ['findmnt', '-o', 'FSTYPE', '-n', '/local4TBraidz/'] failed (): Operation not permitted],

I need to find out what's going on...

anivard

The update on the host are ok.
I reboot the host and the zpool is already mounted.

I try via the XO NEW + STORAGE > ZFS
Same error as before!

message "SR_BACKEND_FAILURE_61(, File SR creation error, )"
name "XapiError"
stack "XapiError: SR_BACKEND_FAILURE_61(, File SR creation error, )\n at Function.wrap (file:///usr/local/lib/node_modules/xo-server/node_modules/xen-api/_XapiError.mjs:16:12)\n at file:///usr/local/lib/node_modules/xo-server/node_modules/xen-api/transports/json-rpc.mjs:38:21\n at runNextTicks (node:internal/process/task_queues:60:5)\n at processImmediate (node:internal/timers:454:9)\n at process.callbackTrampoline (node:internal/async_hooks:130:17)"

olivierlambert

Adding @team-storage in the conversation

anivard

Ok, I am unable to mount the ZFS volume with existing VHD.

I started a new server, I created a EXT STORAGE, I transfered all VHD files.
Good news, all VHD are available in disk section on the EXT STORAGE.

I think, I well recreate the ZFS volume on my primary server and after re-transfert all VHD files.

Unfortunate, it's not possible to mount an already ZFS volume...

All logs are available to understand why it's not possible

anivard

Name:
"API call: sr.createZfs"

Message:
File SR creation error

Stack:
"XapiError: SR_BACKEND_FAILURE_61(, File SR creation error, )\n at Function.wrap (file:///usr/local/lib/node_modules/xo-server/node_modules/xen-api/_XapiError.mjs:16:12)\n at file:///usr/local/lib/node_modules/xo-server/node_modules/xen-api/transports/json-rpc.mjs:38:21\n at runNextTicks (node:internal/process/task_queues:60:5)\n at processImmediate (node:internal/timers:454:9)\n at process.callbackTrampoline (node:internal/async_hooks:130:17)"

anivard

I recreate my ZFS pool.

But next time or for other stuck with this kind of issue, I think you should start to rename your ZFS POOL.
I think the old ZPOOL name stay in XO configuration shadow settings.

To rename a ZFS pool, export it with

zpool export [poolname]

and then import it with the new name using

zpool import [poolname] [newpoolname]

And then recreate (sr-create) your STORAGE in XO