Posts made by tecsolaus

Hello World,

I hope your are healthy and safe in your part of it.

Ive been struggling for a couple of days to resolve this problem. It's essentail to restore one particular VM as it is the sole copy of several years of staff wages data. (There was an unrelated backup failure.)

Original situation:

Xenserver 7.4 running on Fujitsu Primergy server, SRs are on Sysology NAS in high availability (HA) mode with secondary passive NAS acting a failover.
Synology hyperbackup does daily backup to additional NAS.

Objective: Upgrade to xcp-ng.

Method:

Break link between Active and Passive NAS so upgrade is peformed on one data set only, leaving the ability to bring original system back on line if there are problems with the upgrade.

1. Power off primary NAS, and system failed over to secondary as expected.
   The secondary NAS now elevated to be the primary NAS.

2. Shut down all VMs (3 Ubuntu and 2 Win 10 Pro)

3. Do xcp-ng upgrade with Version 8.2 iso installed on USB memory stick.
   All goes well...

4. Start VMs. Linux VMs start ok. Windows 10 VM fails with tapdisk error.

5. I figured at this stage I would bring the old system back online and deal with the upgrade as an isolated system.

6. Shut down NAS with upgraded xcp-ng server.

7. Start (original) primary NAS with original Xen server install.

8. This is where things started going wrong: The upgraded NAS powered up also and started automatically replicating itself to the original primary NAS, overwriting the backup xen server!

9. I removed the HA link between the two NAS units, but too late.

Recovery attempts:

1. Restore VDI volume from backup to one NAS. (16 hr job...)

2. Install xenserver 7.4 on spare machine and connect to NAS volume. All SRs are available but with no details.

3. Export ova file of wanted vm and inport into test server.

4. Experiment by attaching each VDI to SM in turn and booting. (I know the drive I need is 60G which narrows the list to test.)

5. Attach properties labels to each VDI as they are proven to not be the wanted one.

6. I am left with 6 images, all which fail to boot with a tapdisk error. I assume (hope!) the VDI I need is one of these.

This is one failure example from: [root@xen74 log]# grep -i -r "tapdisk"
xensource.log:

Jan  7 08:20:01 xen74 xenopsd-xc: [debug|xen74|32 ||xenops_server] TASK.signal 178 = ["Failed", ["Storage_backend_error", ["SR_BACKEND_FAILURE_453", ["", "tapdisk experienced an error [opterr=No such file or directory]", ""]]]]

Jan  7 08:20:02 xen74 xenopsd-xc: [debug|xen74|33 ||xenops_server] TASK.signal 168 = ["Failed", ["Storage_backend_error", ["SR_BACKEND_FAILURE_453", ["", "tapdisk experienced an error [opterr=No such file or directory]", ""]]]]

Jan  7 08:20:02 xen74 xapi: [error|xen74|15251 |Async.VM.start R:71526f06b73d|xenops] Re-raising as SR_BACKEND_FAILURE_453 [ ; tapdisk experienced an error [opterr=No such file or directory];  ]

Jan  7 08:20:02 xen74 xapi: [error|xen74|15251 ||backtrace] Async.VM.start R:71526f06b73d failed with exception Server_error(SR_BACKEND_FAILURE_453, [ ; tapdisk experienced an error [opterr=No such file or directory];  ])

Jan  7 08:20:02 xen74 xapi: [error|xen74|15251 ||backtrace] Raised Server_error(SR_BACKEND_FAILURE_453, [ ; tapdisk experienced an error [opterr=No such file or directory];  ])

I've experimented with adding and removing CD drives as this was mentioned in numerous forums. No difference.

I'm stuck, way out of my depth, and looking for any advice which might assist! I'm happy to pay for professional support, but the time frame looks like it could be quite long so I am contacting the community forum as well.

Cheers, James Dean