BRAND_CONSOLE error on BOOT, no networking

andremartin

Hey guys!

So we had a power outage here that shutdown our xcp-ng server. Upon returning, it goes through most of the boot process, it shows the splash screen, but then it gets stuck on this error:

The host has no network connectivity.
Going to another console with alt+right, I saw that emergency mode is "false", and tried network restart, which made no difference. Boots up to the same error.

Tried a toolstak restart, and this is what xensource.log looks after that:

xensource.log.jpg

And this is what /etc/xensource-inventory looks like:

Sorry for the print screens. My only access is either the KVM or the iDRAC console which doesn't let me copy anything (afaik).

Any ideas? Did something got corrupted? If it did, can it be rebuilt?

I do have access to USB ports to do a re-install, but well, I would rather not have to redo the entire server.

stormi

/etc/xensource-inventory is supposed to have a lot more contents. Here's mine from a test XCP-ng VM:

[10:23 xcp-ng-sve-80-dev ~]# cat /etc/xensource-inventory
PRIMARY_DISK='/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
PRODUCT_VERSION='8.0.0'
DOM0_VCPUS='2'
CONTROL_DOMAIN_UUID='585098b9-5c1e-49d7-875c-e88c105b6ac1'
DOM0_MEM='1136'
COMPANY_NAME_SHORT='Open Source'
MANAGEMENT_ADDRESS_TYPE='IPv4'
PARTITION_LAYOUT='ROOT,BACKUP,LOG,BOOT,SWAP,SR'
PRODUCT_VERSION_TEXT='8.0'
PRODUCT_BRAND='XCP-ng'
INSTALLATION_UUID='fc51eb64-b541-4a1a-8cea-2bc39590be81'
PRODUCT_VERSION_TEXT_SHORT='8.0'
BRAND_CONSOLE='XCP-ng Center'
PRODUCT_NAME='xenenterprise'
MANAGEMENT_INTERFACE='xenbr0'
COMPANY_PRODUCT_BRAND='XCP-ng'
PLATFORM_VERSION='3.0.0'
BUILD_NUMBER='release/naples/master/45'
STUNNEL_LEGACY='false'
PLATFORM_NAME='XCP'
BRAND_CONSOLE_URL='https://xcp-ng.org'
BACKUP_PARTITION='/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001-part2'
INSTALLATION_DATE='2019-08-06 15:52:19.379300'
COMPANY_NAME='Open Source'

andremartin

So, is there a "proper" way to rebuild this?

stormi

It is created during installation, by this python code: https://github.com/xcp-ng/host-installer/blob/eb796dc6a82a0defdcbba1a0833a3356060c5396/backend.py#L1625

I don't know if there's a way to re-create it.

andremartin

Loaded the 8.0 Install ISO to see if there was some recovery option, but it doesn't even recognize that there's an installation on the disk...

stormi

@andremartin Yeah, it recognizes existing installations... Based on /etc/xensource-inventory if I remember correctly...

stormi

If you had upgraded that host in the past, you should be able to find an /etc/xensource-inventory in the backup partition, you could then use after changing the values that depend on the version.

stormi

Another place where you can find a backup, if you have made one before the crash: in the backup file that the xe host-backup command creates. It's also available from XSConsole. XCP-ng Center also offers that feature if I remember correctly.

andremartin

I do have a backup partition, but I can't seem to be able to mount it? The filesystem shows up as "Microsoft basic". Which filesystem should I specify to mount it?

stormi

It's ext3, unless it's empty. It only gets filled when you upgrade, so that you can fallback to the previous state if anything goes wrong.

andremartin

Then I guess it's empty. I guess I'll just redo the server, and chalk this up as a one in a million file corruption, for which we didn't have the proper OS backup to counteract. We do have backups of the VMs anyway.

Thanks for the help! We didn't get this fixed, but I did learn a lot! This is our first Xenserver/XCP host.