#NEW: NAbox 4.0.11 has been released
1 messages · Page 1 of 1 (latest)
during upgrade, getting stuck at the reboot
[ 3.833595] EXT4-fs (sda9): mounted filesystem 84a5cefa-c3c7-47d7-9305-7e6877f73628 r/w with ordered data mode. Quota mode: none.
[ 3.868979] BTRFS info (device sda6): first mount of filesystem 50e7422b-f0c7-4536-902a-3ab4c864240b
[ 3.874177] BTRFS info (device sda6): using crc32c (crc32c-intel) checksum algorithm
[ 3.878411] BTRFS info (device sda6): using free space tree
[ 3.880192] BTRFS info (device sda6): auto enabling async discard
[ 4.039205] BTRFS info (device sda6): last unmount of filesystem 50e7422b-f0c7-4536-902a-3ab4c864240b
[ 4.084257] BTRFS info (device dm-0): last unmount of filesystem 88a9bed3-d26b-40c9-82ba-dbb7d44acae7
[ 4.086203] EXT4-fs (sda9): unmounting filesystem 84a5cefa-c3c7-47d7-9305-7e6877f73628.
after a while it reboots on its own but sticks at the same place and goes into a boot loop
What was the previous version ?
4.0.10 f9baaca
not sure if it matters or not, but I do have an NFS share attached to the previous version that was for the data
i can send the entire boot log over from the serial console if you want
that would be it, not sure why though
or why it would prevent booting
yes please send the logs
Well… that’s Hyper-V right ?
yes
I don't see anything wrong with hyper-v packaging of the update file. Do you have a way to restore a snapshot and log the upgrade ?
dont believe so, i didnt even think about taking a snapshot.
anyway to do a boot with just bare minimum enabled, no nfs/etc or anything
let me go look, maybe an older snapshot is around
hrm. only one available is from 12/26 last year, so doubt it would be useful
the nfs export has some logs/etc
update.json
{"percent":75,"title":"Rebooting","description":"Rebooting","currentStage":"stage2","doneStages":["stage1"],"finished":false,"started":true,"skipOSUpgrade":false}
about to head out, will restore the snapshot in a couple of hours when i start working again
I think serial dump is missing informations actually sent to screen, not entirely sure though, if you can take a look at actual screen if something more helful is in there
I'm comparing a vmware boot with the serial Hyper-V boot (couldn't figure out if VMware flatcar has serial enabled, it doesn't look like it) and I can definitely make up things similar formatted as raw output, and some additional logs on the vmware side on screen
anyway to enable something to log via the boot menu?
Not that I'm aware 🙁 I was actually screen recording the vmware boot 😄
that's the only other thing that gets show in the serial console, then it does "restarting system" and reboots, over and over
serial console.... the actual console of the system, whatever you call it but that's what azure calls it
after i take a nap and start back to work i'll export the drive and put it on a local hyper-v machine, see what it does.
I'll also push out a new system with that snapshot and do the upgrade from there, as well as a new build with the hope of being able to restore the data from the old system if all else fails.
thanks, i'll be back in a couple of hours
Hello, updated yesterday, so far so good.
Same here, have seen no issues after update.
Ah it’s not only hyper-v but azure.
If there is no error maybe it’s stuck upgrading forever ? You’re never getting to the login prompt or ssh availability ?
correct, it goes through that startup cycle i sent you over and over
@median basin uprade from 4.0.10 to 4.0.11 worked smooth in our VMware environment 😊 very good job with the new version specially the easy integration of the switches
have not been able to get anything to boot back into this, so I'm going to build a new instance.
Hopefully there is a way I can pull the data and config over from the old install?
We can figure this out, you can attach de data drive and we can check what's available
Hi @median basin
I’m planning to migrate the NAbox 3 to 4 ; I have question -
- if we migrate the NAbox then grafana address will also change and whoever is using old grafana address should be moved to new address
- what if even manual migration fails - can we roll back
Thank you
- you could keep the IP address when migration is done for minimal disruption. There might be some changes related to harvest dashboards evolving but chances are it’ll be fine.
- since it’s an actual migration, the old appliance can keep running and doesn’t skip a beat, so yes, people still use nabox3 until you’re all set with NAbox 4
Thanks a lot Yann
As I understand after migration all will be migrated too NAbox4 but we can keep using both as we will have different IPs , that looks nice as we can have some issues post migration on NAbox4 grafana dashboard , till the time I will be all good with NAbox4 I wil keep using 3
The update worked like a charm, thank you so much for this wonderful tool! I especially like the new feature to monitor the switches from the Metrocluster.
So as I understand it will only monitor CISCO switch
- metro cluster
- inter cluster switch
- fabric switch - not supported
Can you please help me with the documents to add settings on the switch
Hi @void moat did you see this documentation? https://netapp.github.io/harvest/nightly/prepare-cisco-switch/
Thanks Chris
@median basin Hey, coming back around to this since the failed upgrade.
i have a new install setup, just need to see if we can migrate the data from the old disks to the new one without losing all of it's data or configs
You were using nfs right ?
yes
So can you just reattach the mount ? Or you won't be using NFS now ?
i can still use NFS, but will that bring over the old config/etc
Ok, for the config you can just copy harvest.yml file, and meybe the .env.custom file
alrighty.
got the 2 files copied over, but can't log in to grafana and all the settings for network, ldap, ssl and email are gone
so, the whole nabox directory should be fine. I'm just not sure what could have been changed and was hesitant to tell you to move everything over. SSL is in /etc/nabox/secrets/ssl, ldap is in /etc/nabox/grafana/ldap.toml, network is /etc/systemd/network/10-nabox.network and email is /etc/nabox/alertmanager/alertmanager.yml